sample literature report - personal.tcu.edupersonal.tcu.edu/jmontchamp/f2019 files for...

SAMPLE LITERATURE REPORTLITERATURE PROBLEM for FLUTICASONE PROPIONATE

Class ID #: 007

1. Structural Formula: 2. Compound Name

Drawn using ChemDraw Prime 15.1

3. Molecular Formula: 𝐶"#𝐻%&𝐹%𝑂#𝑆 4. Molecular Weight: 500.57

5. Physical Properties:

Pure, white crystalline solid. Melting Point: 274-275° C 1

6. Spectral Properties:

Infrared: Recorded in the range from 650-4000 cm-1 at 4 cm-1 spectral resolution: 3336 cm-1 (m, δ, O-H) 3069 cm-1 (vw, n, C-H) 3051 cm-1 (vw, n, C-H) 3024 cm-1(vw, n, C-H of five-membered ring) 2975 cm-1 (m, n, C=C-H) 2964 cm-1 (m, nas, CH3) 2942 cm-1 (m, nas, CH2) 2881 cm-1 (m, ns, CH2) 1774 cm-1 (s, n, C-O) 1699 cm-1

(s, n, S-C-O) 1661 cm-1 (vs, n, C-O) 1616 cm-1 (s, n, C=C) 1530 cm-1 (vw, δ, CH3) 1455 cm-1 (m, δ, CH3) 1395cm-1 (mw, δ, CH2) 1365 cm-1 (mw, δ, C-H) 1304 cm-1 (m, n, C-C) 1271 cm-1 (m, n, C-F) 1027 cm-1 (m, n, F-C-S) 993 cm-1 (s, r, CH3) 782 cm-1 (w, r, CH2). 2

Proton NMR: 1H (90 MHz) 0.96 (16a-CH3, d, J = 7Hz), 1.06 (propionate CH3, t, J = 7 Hz), 1.07 (13-CH3, s), 1.56 (10-CH3, s), 2.33 (propionate CH2, q, J = 7 Hz), 4.32 (11a-H, broad m), 5.63 (11-OH, broad d, J = 4 Hz), ca. 5.75 (6-H, broad dm, J = 50 Hz), 6.00 (SCH2F, d, J = 51 Hz), 6.19 (4-H, m), 6.37 (2-H, dd, J =10, 2 Hz), 7.33 (1-H, d, J = 10Hz). 1

Carbon NMR: 13C (101 MHz, DMSO-d6 ) δ 192.87 (s), 183.98 (s), 172.07 (s), 162.43 (d, J = 13.5 Hz), 151.54 (s), 128.91 (s), 119.25 (d, J = 12.1 Hz), 99.72 (d, J = 176.3 Hz), 95.94 (s), 86.62 (d, J = 178.0 Hz), 80.92 (d, J = 211.8 Hz), 69.97 (d, J = 37.2 Hz), 48.40 (s), 47.84 (d, J = 22.4 Hz), 42.84 (s), 35.74 (s), 35.10 (s), 33.73 (d, J = 19.4 Hz), 33.37 (s), 31.93 (m), 26.94 (s), 22.73 (s), 16.95 (s), 16.08 (s), 9.05 (s). 3

Fluorine NMR: 19F (376 MHz, CDCl3 ) δ − 165.35 (dd, J = 27.5, 8.5 Hz), − 187.00 (dd, J = 48.3, 13.8 Hz), − 191.35 (t, J = 49.6 Hz). 3

Mass Spectra: MS (ESI+) m/z 501.0 [M + H]+.3

Flonase (commercial name)Fluticasone Propionate (activecompound)

systematic nomenclature?

7. SYNTHETIC ROUTE:

First compound is flumethasone, which is referred to in Kertesz as being one of four baseline “commercially available” steroid hormones that one can make thiol esters from.4 Given that Flonase is sold for around 10-15$ for a total of 3000 mcg of active compound (0.003 grams), 1 gram of the active compound would provide ~333 bottles or 20,000 doses of the drug. Thus, 1 gram would equal, in a low estimate, about $3,300. Thus, I don’t think it’s unreasonable that the previously described “commercially available” precursor, flumethasone is priced at $1042 per gram, given as we’re essentially producing, considering yields and lowballing, the active compound for 10,000 doses of a drug. 8. SHORT SUMMARY Fluticasone Propionate (FP) is an androstane carbothioate glucocorticosteroid that is prescribed to treat inflammatory respiratory conditions such as asthma and allergic rhinitis. As opposed to other glucocorticosteroids, FP has a very low oral bioavailability and is measured to have approximately 100 times more concentration during its half-life in peripheral and 300-400 times higher in central lung tissue as opposed to serum. 5 It is also retained in lung tissue and has been found to saturate glucocorticosteroid receptors, binding with nearly full affinity.6 For medicinal purposes, it is most commonly administered as a nasal spray with dosing from 50-2000µg per day in adults.5

I like the reasoning and argumentation, but it still would have been wise to find where the starting material comes from

9. BIOLOGICAL MODE OF ACTION Fluticasone propionate is a glucocorticosteroid, and in this class of compounds, the main effect is a reduction of inflammation. This is generally achieved by many kinds of glucocorticoids by inhibiting the expression of various components of the immune and inflammatory systems such as cytokines, chemokines, cell-surface glucocorticoid receptors, adhesion molecules, tissue factor, degradative proteinases, and certain enzymes that produce inflammatory mediators.7 When considering the specific example of glucocorticoids on lung tissue, the immediate decrease in inflammation seems to be caused by the reduction of mast cells (derived from the German Mastzellen, which means “well-fed cells” due to their high concentration of granules in their cytoplasm). Mast cells contain many inflammatory mediators such as histamine, proteases, and cytokines.8 The reduction in mast cells seems to be caused by the subsequent reduction of IL-3, which is necessary for mast cells to survive.9 Fluticasone propionate has also been shown to reduce levels of IL-410, but this would be indicative of the reduced number mast cells not releasing IL-4, as the release of inflammatory mediators does not seem to be decreased. Therefore, it is due to the decrease in the number of mast cells based on IL-3 that is a more likely model. As stated on the drug facts information for fluticasone propionate, the “the precise mechanism through which fluticasone propionate affects allergic rhinitis symptoms is unknown”, but there are several models for how glucocorticoids reduce the production of inflammatory factors, mostly based on transcriptional interference. The mechanism for this would be that the glucocorticoid would bind to the glucocorticoid receptor, which then would directly bind to two transcription factors (activator protein-1 and nuclear factor kB), thus leaving them unable to bind to the DNA as a transcription factor and produce the associated protein for an inflammatory mediator (interleukins, cytokines, etc.). A second proposed mechanism is that the liganded glucocorticoid receptor inhibits the MAP kinase pathway by keeping the kinases dephosphorylated.7 In more recent literature, these theories have continued to be supported, with the added discovery that there are different types of glucocorticoid receptor isoforms and that a given glucocorticoid will confer a different conformational change in the receptor that will then change either the binding of transcriptional factors to the DNA, effect a change on the chromatin of the strand, or influence a co-receptor.11 In terms of fluticasone propionate, the discovery that the glucocorticoid receptor binds to NF-kB at relatively low glucocorticoid concentrations compared to the requirement to induce gene expression, or otherwise put, it is more likely that the mechanism occurs by repressing the expression of inflammatory mediators rather than inducing the expression of inflammatory inhibitors.12 10. BIBLIOGRAPHY

1. Phillipps, G.H; Bailey, E. J.; Bain, B.M. et al. Journal of Medicinal Chemistry. 1994. 34, 3717. <http://pubs.acs.org/doi/pdf/10.1021/jm00048a008> Accessed on 11/30/16.

2. Ali, H.R.H.; Edwards, H.G.M.; Kendrick, J.; Scowen, I.J.; Spectrochemica Acta: Molecular and Biomolecular Spectroscopy. 2009. 72 (2), 244. <http://dx.doi.org.ezproxy.tcu.edu/10.1016/j.saa.2008.08.004>. Accessed on 11/18/16, 11/30/16.

3. Zhao, J; Jin, C.; Weike, S. Organic Process Research & Development. <http://anothersample.net/improved-synthesis-of-fluticasone-propionate> Accessed on 11/28/16.

4. Kertesz, D.; Marx, M. J. Org. Chem. 1986. 51: 2315. <http://pubs.acs.org.ezproxy.tcu.edu/doi/abs/10.1021/jo00362a028> Accessed on 11/28/16.

5. Holliday, S.M.; Faulds, D.; Sorkin, E.M. Drugs. 1994. 47: 318. <http://link.springer.com/article/10.2165/00003495-199447020-00007> Accessed on: 11/20/16.

6. Esmailpour, N.; Högger, P.; Rabe, K. F.; Heitmann, U.; Nakashima, M.; Rohdewald, P. European Respiratory Journal. 1997. 10:1496-1499. <http://erj.ersjournals.com/content/erj/10/7/1496.full.pdf> Accessed on: 12/1/16.

7. Saklatvala, J. Arthritis Research & Therapy. 2002. 4:146. <https://www.ncbi.nlm.nih.gov/pmc/articles/PMC128923/pdf/ar398.pdf> Accessed on: 12/4/16

8. Amin, K. Respiratory Medicine. 2012. 106: 9. <https://www.ncbi.nlm.nih.gov/pmc/articles/PMC128923/pdf/ar398.pdf> Accessed on: 12/4/16

9. Van der Velden, V.H.J. Mediators of Inflammation. 1998. 7:229.<https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1781857/pdf/9792333.pdf> Accessed on: 12/4/16.

10. Masuyama, K.; Jacobson, M. R.; Rak, S. et al. Immunology. 1994. 82:192.<https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1414820/pdf/immunology00081-0026.pdf>Accessed on: 12/4/16

11. Oakley, R.H.; Cidlowski, J.A. J Allergy Clin Immunol. 2013;132(5):1033.<http://ezproxy.tcu.edu/docview/1504841368?accountid=7090> Accessed on: 12/4/16.

12. Adcock, I. M.; Nasuhara, Y.; Stevens, D.; Barnes, J. British Journal of Pharmacology. 1999. 127:1003. <http://onlinelibrary.wiley.com/doi/10.1038/sj.bjp.0702613/pdf> Accessed on: 12/4/16.

REMEMBER THAT THE LITERATURE REFERENCES EMPLOYED MUST BE PROVIDED. (can be as email attachments)

HOWEVER, IF YOU PREFER, YOU MAY ONLY PROVIDE THE

RELEVANT PAGE.

IN THIS SAMPLE REPORT, ALL THEREFERENCES ARE PROVIDED AND

THE DATA IS HIGHLIGHTED

J. Med. Chem. 1994,37, 3717-3729 3717

Synthesis and Structure-Activity Relationships in a Series of Antiinflammatory Corticosteroid Analogues, Halomethyl Androstane-l7/3-~arbothioates and - 17fl-carboselenoates

Gordon H. Phillipps, Esm6 J. Bailey, Brian M. Bain,* Raymond A. Borella, Jacky B. Buckton, John C. Clark, Alice E. Doherty, Alan F. English, Harold Fazakerley, Stuart B. Laing, Elizabeth Lane-Allman, John D. Robinson, Peter E. Sandford, Peter J. Sharratt, Ian P. Steeples, Robert D. Stonehouse, and Christopher Williamson Glaxo Research and Development Limited, Greenford Road, Greenford, Middlesex UB6 OHE, U.K.

Received September 20, 1993@

The preparation and topical antiinflammatory potencies of a series of halomethyl 17a-(acyloxy)- 11~-hydroxy-3-oxoandrosta-1,4-diene-17~-carbothioates7 carrying combinations of 6a-fluoro, 9a- fluoro, 16-methy1, and 16-methylene substituents, are described. Key synthetic stages were the preparation of carbothioic acids and their reaction with dihalomethanes. The carbothioic acids were formed from 178-carboxylic acids by initial reaction with dimethylthiocarbamoyl chloride followed by aminolysis of the resulting rearranged mixed anhydride with diethylamine, or by carboxyl activation with 1,l’-carbonyldiimidazole (CDI) or 2-fluoro-N-methylpyridinium tosylate (FMPT) and reaction with hydrogen sulfide, the choice of reagent being governed by the 17a-substituent. Carboxyl activation with FMPT and reaction with sodium hydrogen selenide led to the halomethyl 16-methyleneandrostane-17~-carboselenoate analogues. Anti- inflammatory potencies were measured in humans using the vasoconstriction assay and in rats and mice by a modification the Tonelli croton oil ear assay. Best activities were shown by fluoromethyl and chloromethyl carbothioates with a 17a-propionyloxy group. 5’-Fluoromethyl 6a79a-difluoro-l l~-hydroxy-l6a-methyl-3-oxo-l7a-(propionyloxy)an~osta-l,4-diene-l7~-carbothioate (fluticasone propionate, FP) was selected for clinical study as it showed high topical antiinflammatory activity but caused little hypothalamic-pituitary-adrenal suppression after topical or oral administration to rodents.

The successful development of corticosteroid analogues designed to show high potency on local application to inflamed tissue has been Although the avilable compounds showed only weak undesirable systemic side effects after topical administration, we continued to seek further improvement. In this paper we describe a series of potent and novel corticosteroidal halomethyl esters of androstane-l7#?-carbothioic acids with promising separations of a~ t iv i ty .~

Earlier we reported that the normal two-carbon 178- side chain of pregnanes was not necessary for corticoid activity, androstane-l7~-carboxylates showing high topical antiinflammatory activity if both the l7a-hydroxy and 176-carboxylic acid functions were esterified, with the greatest activity being shown by l7a-propionates as fluoromethyl carboxylates.116-8

We first explored the synthesis of l7a-unsubstituted 17B-carbothioates in making compounds with anaes- thetic a~tivity.~JO Kertesz and Marxll overcame many difficulties while employing carboxyl-activation procedures followed by reactions with alkanethiols in the synthesis of some alkyl l7a-acyloxy 17B-carbothioates which showed good topical antiinflammatory activity. Their synthetic methods could not, however, be applied to the preparation of halomethyl carbothioates (analogues of the potent 21-halopregnan-20-0nesl-~) as halomethanethiols are not known and would be expected to be very unstable. We therefore devised new methods of preparing 17a-hydroxy and l7a-acyloxy 17j3-

~~~

* External Scientific Affairs Department, to whom communications

e Abstract published in Advance ACS Abstracts, September 15,1994. should be sent.

carbothioic acids, which could then be esterified by reacting their salts with dihalomethanes. Furthermore, l7a-acylation of haloalkyl l7a-hydroxy 17p-CarbOthiO- ates was likely to be unselective in the presence of an llp-01, so it would be preferable to l7a-acylate at the carbothioic acid stage, as with the carboxylic acid^.^$^ The antiinflammatory potencies of the halomethyl 17a- acyloxy 17/?-carbothioates were in general high, as delineated in Table 3. ’

Chemistry The 17a-hydroxyandrostane-17fi-carboxylic acid in-

termediates 2 for the desired carbothioates were prepared by oxidative cleavage of 21-hydroxypregnan-20- ones 1 with periodic acid in aqueous dioxane or THF. They were readily 17a-acylated, without concomitant lip-acylation, by reaction with excess acyl chloride and triethylamine followed by aminolysis of the resulting l7a-acyloxy 17/3-carboxylic acid mixed anhydrides with diethylamine1J’v8 to give the corresponding acids 3 (Scheme 1).

Our f i s t synthesis of carbothioic acids resulted while studying the aminolysis of the mixed anhydride formed by reaction of the l7a-propionyloxy l7fi-carboxylic acid 3dll with dimethylthiocarbamoyl chloride in pyridine. The mixed anhydride was initially believed to be the thione 4a, but aminolysis with diethylamine gave the l7p-carbothioic acid 6i and it now has been assigned the rearranged structure 4 b ~ ; ~ this was supported by its synthesis as the major product from the reaction of thioacid 6i with dimethylcarbamoyl chloride and by the lack of a carbonyl IR absorption above 1740 cm-l, in contrast to that observed (1783 cm-l) for the oxygen

0022-2623/04/l037-3717$04.50l0 0 1994 American Chemical Society

3718 Journal of Medicinal Chemistry, 1994, Vol. 37, No. 22 Phillipps et al.

Scheme 1

Scheme 2

C0,H CO-H

1 2

4a x=o, Y=S 46 XIS, Y=O

COSH

5

Scheme 3

- OCOC,H, M e

CM 0 & / 3d

analogue 4ca (see Scheme 2, also Table 1 for the definition of these compound letters).

Gais12 reported that carboxylic acid imidazolides are accessible in almost quantitative yields from carboxylic acids and 1,l'-carbonyldiimidazole (CDI) and that they react rapidly with alphatic and aromatic thiols to give thioesters in high yields. Kertesz and M a d 1 applied this method to the synthesis of alkyl androstane-l7/I- carbothioates. We found that the imidazolides from l7a-hydroxy 17/I-carboxylic acids also react with hydrogen sulfide to give l7a-hydroxy 17~-carbothioic acids 6. Like the corresponding carboxylic acids, these were readily and selectively l7a-acylated without concomitant 11/3-acylation (Scheme 2).

Kertesz and M a d l reported that CDI failed to activate l7a-acyloxy 17/I-carboxylic acids, but we found that CDI in dimethylformamide at 22 "C activated the

3

4b

7 8

1-

17-propionate acid 3d to reaction with NaSH (to give 6i) or NaSMe (to give 9) but not to reaction with H2S. The reaction of H2S with other carboxyl-activated species is known to be base-~ata1yzed.l~ The known thioester Sll could also be prepared by methylation of 6i. The nature of the activated species was studied by partitioning the reaction mixture between water (not acid) and ethyl acetate (Scheme 3). The two major components isolated by preparative TLC were the diastereoisomeric 17-spiro-2'-( l-imidazolyl)-l',3'-diox- olan-4'-0nes 7. The structures of each were indicated by their characteristic IR absorption (1810 cm-l) for the spirocyclic carbonyl and by their 13C NMR spectra, where the spirocyclic ring carbon atoms were readily assigned. In particular the 2'-carbon resonances (112.1 and 112.0 ppm) in the two isomers showed no C-H coupling in the off-resonance spectra. These and the

Antiinflammatory Corticosteroid Analogues

Table 1. Physical Properties of Intermediates"

Journal of Medicinal Chemistry, 1994, Vol. 37, No. 22 3719

[a111 deg UV, (EtOH) mp, "C (dioxane) formula ana1.b nm ( E x no. 6a 9a 16 17a 17P

2a 2b 2c 3a 4ba 4bb 4bc 4bd 4ca 5a 5b 5c 5d 5e 6a 6b 6c 6d 6e 6f 6g 6h 6i 6j 6k 7a 7b 8 1Oi lOj 10k 101 l l b l l c l l d l l e l l f 1lg l l h 16 17 18a 18b 19a 19b 20ad 20bd 21ad 21bd 22c 22d

F H H F F F F H H F H F H F H F H F H F H F F F H F F F F H H F H F F F F F F F H H H F H F H F H F H F H F H F F H H F F F F F F H H F F F F F F F H F H F H F H F H F H F H F H F H F H F H F H F H F H F

OH COzH OH COzH OH COzH OCOCzHs COzH OCOCqHn COSCONMez OCOCk3- COSCONMei OCOCzH5 COSCONMez OCOCzH5 COSCONMez OCOCzH5 COzCONMez OH COSH OH COSH OH COSH OH COSH OH COSH OCOCzHs COSH OCOCzH5 COSH OCOCzH5 COSH OCOCH3 COSH OCOCzH5 COSH OCOC3H7 COSH OCOCzH5 COSH OCOCH3 COSH OCOCzH5 COSH OCOC3H7 COSH OCOCzH5 COSH

spiro(CgHgN~03Y spiro(CgH&03Y

OCOCzH5 CO(C3H3Nz) OCOCzHs COSCHzCl OCOCzH5 COSCHzCl OCOCH3 COSCHzCl OCOC3H7 COSCHzCl OCOCzH5 COSCHzI OCOCzH5 COSCHzI OCOCH3 COSCHzI OCOCzH5 COSCHzI OCOC3H7 COSCHzI OCOCH3 COSCHzI OCOCzHs COSCHzI epoxide COSCHzCl OH COSCHzCl epoxide COSeCH3 epoxide COSeCHzCl OH COSeCH3 OH COSeCHzCl OH COSeCH3 OH COSeCHzCl OCOCzH5 COSeCH3 OCOCzH5 COSeCHzCl OCOCzH5 COSeCHzI OCOCZH~ COSeCHzF

24 H F &CHi OCOC,H, COSlf

24 1-248 289-293 248-252' 224-227 191-193

167-170 189-192 200-203 222-225 209-214 230-232 198-201' 251-256' 189-193 136-138 141-143 175-177 161-164 155-157 159-163 178.5-179 177-179 175-176 236-239 193-195 168-176 190-203 188-191 247-250 280-283 235-238 195-197 175-178 241-243 233-236' 210-212 204-205 191-199 246-251 242-243 297-299 229-232' 248-254' 237-239 225-235' 209-210 209- 2 11 156-158 174- 175 216-218 234-235

foam

+54 +66q -24 +3

+82 +172 +185

+6 +74

+116 $104 +94

+189 +58 $72 $75 $30 -10 -27 +21

1113 +98

+110 +lo7

-71 $96 -8

f 5 2 $48 +50.5 +45 +49 $18 +4

+78 +87 +89 -29 -31

+131 +36

+156 +111 +82.5 +45.5

+124 +142

-9 -8

-70 -43.5

+lo7

242 (15.3) 239.5 (14.5) 238 (16.0) 242 (14.3) 239 (18.7) 238 (19.6) 237.5 (19.8) 236.5 (19.2) 238 (15.9) 244 (19.3) 245 (17.6) 243 (21.0) 243 (19.5) 242 (20.5) 247 (18.1) 245 (18.4) 243 (16.3) 242.5 (18.0) 243 (18.9) 242.5 (18.4) 247 (17.8) 242 (17.8) 242 (18.3) 244 (18.1) 242 (17.7) 240 (14.3) 240 (14.5) 240 (17.3) 240 (18.7) 238 (19.0) 238.5 (19.8) 238.5 (19.9) 243.5 (20.7) 242 (17.9) 240 (20.2) 241 (20.6) 241 (20.7) 241 (20.6) 241 (19.8) 239 (20.5) 239.5 (19.4) 239 (16.3) 237.5 (16.5) 240.5 (15.9) 240.5 (16.3) 238 (15.7) 237 (16.1) 237 (16.0) 237 (16.1) 238 (16.7) 239.5 (16.8) 238 (37.1)

~~

a lH NMR and infrared spectra were obtained for all compounds, and data are in the Experimental Section for selected compounds. b Persistent solvation was confirmed spectroscopically. ' Decomposition. 11-Ketones. e Solvate (0.5EtOAc). f Solvate (0.25MezCO). g Solvate (1.OHzO). Solvate (0.25H20). Isomer A. J Isomer B. Solvate (0.16EtOAc). Solvate (0.2EtOAc). Solvate (0.5HzO). Solvate (0.33H20).

Solvate (0.6MeOH). p Disulfide.

C-17 resonances (92.1 and 91.1 ppm) were structurally characteristic, and the lH NMR confirmed tha t the ethyl groups in each isomer were no longer par t of a n ester function; the configurations of the two isomers were not established. From a further experiment, in which the reaction mixture was t reated with dimethylamine, the same dioxolanones 7 were isolated together with a minor third component of intermediate polarity, identified a s the imidazolide 8, Y,, 1742 cm-l. The dioxol-

anones 7 each reacted with NaSMe to give 9, but not with NaSH. There was insufficient 8 to tes t its reactivity.

Chloromethyl carbothioates 10 were prepared from carbothioate salts by alkylation with bromochloromethane, or chloroiodomethane, in dimethylacetamide. The chloromethyl thioesters 10 reacted with sodium iodide to give iodomethyl thioesters 11, which in t u r n reacted with sodium bromide or silver fluoride


Table 2. Biologically-Assayed Halomethyl Androstane-17j3-carbothioates and -17j3-carboselenoates (Physical Propertiesa)

f i 0 ' ?

Y

COSeCHg H o ~ y COSCH,X @OCOR

16

0 Y

10,11,12,13 22

no. Z Y X R 16 mp, "C [a l~, detg (dioxane) formula anal. UV, Amax (EtOH) nm ( E x

13a H F F CzH5 H 224-225 +70 CZ~H~OFZOSS C,H,F,S 238.5 (17.4) 10a F H C1 CZH6 H 196-199 +38 Cz4H30ClF05S C,H,S 238 (16.9) 13b F H F CzH5 H 207-211 +70 C Z ~ H ~ O F Z O ~ S C,H,F,S 237 (17.1)

+37 C Z ~ H ~ Z F Z O ~ S C,HS 237 (18.0) 13c F H F CzHs a-CH3 242-243 lob F F C1 C Z H ~ a-CH3 272-275 +49 Cz5H31ClFzOsS C,H 238 (20.0) 13d F F F CH3 a-CH3 308-310 +29 Cz4HzgF305S C,H,S 236 (19.0) 13e F F F CzH5 a-CH3 274-275 +32 Cz~H3iF305S C,H,F,S 236.5 (18.6) 13f F F F C3H7 a-CH3 249-252 +32 Cz6H33F305S C,H,F,S 237 (19.1) 1 0 ~ H H C1 CzH5 j3-CH3 192-193 +65 Cz5H33C105S C,H,Cl,S 241 (17.8) 13g H H F CzH5 j3-CH3 223-225 +lo3 Cz5H33F05S C,H,S 240 (16.3) 10d F H C1 CH3 j3-CH3 220-223 f39.5 Cz4H3oClFOsS C,H,Cl,S 238 (18.4) 13hb F H F CH3 j3-CH3 248-249 +lo1 CZ~H~OFZO~S C,H,F,S 237 (17.7) 10e F H C1 C2& j3-CH3 212-214 +44 C Z ~ H ~ Z C ~ F O ~ S C,H,Cl,S 238.5 (18.8) 1% F H Br CZH5 j3-CH3 186.5-187 +2 Cz&zBrFOsS C,H,Br,S 241 (19.9) l lac F H I CzH5 j3-CH3 196-197 -32 C Z ~ H ~ ~ F I O ~ S C,H,I,S 242 (20.2) 13i F H F CzH5 j3-CH3 237-241 +98 CZ~H~ZFZOSS C,H,F,S 237 (17.4) 10f F H C1 C3H7 j3-CH3 172-175 +46 CzaH3&lF05S C,H,Cl,S 239 (18.7) log F H C1 CZH5 =CH2 212-221 -56 CzsH30ClF05S C,H,Cl,S 239 (19.5) 13j F H F CzH5 =CHz 205-215 -58 CZ~H~OFZOSS C,H,FS 237 (18.1) 10h F F C1 CzH5 =CH2 242-245 -56 CzsHzgClFzO5S C,H,Cl,S 238.5 (19.9) 13k F F F CzH5 =CHz 2f11-255~ -56 Cz5HzgF305S C,H,S 236.5 (19.1) 22a F H H CZH5 =CHZ 225-227 -37 CzsH31F05Se C,H,Se 239.5 (16.3) 22b F H C1 CzH5 =CHz 212-214 -48 Cz~H30ClF05Se C,H,Cl,Se 240.5 (16.6)

* Solvated with 0.5 mol of HzO. Solvated with 0.33 mol of HzO. Decomposition.

Scheme 4

a 'H NMR and infrared spectra were obtained for all compounds, and data are in the Experimental Section for selected compounds.

OT BrCHaCI W p p):z I - [O

{S [fi ;

6 ____) NoHCO,

'iBr- \ ~ Q F

12 13

BrCH,F or ICH,F, base

t o give the bromomethyl and fluoromethyl thioesters 12 and 13, respectively. The fluoromethyl thioester 13e was also prepared directly from the potassium salt of the carbothioic acid 6e, using fluor~iodomethanel~ or bromofluoromethanel5 (Scheme 4).

Kertesz and M a d l reported that 2-fluoro-N-methylpyridinium tosylate (FMPT) could be used for the activation of 17,+carboxylic acids in the presence of a 16aJ7a-acetonide but that neighboring group participation dominated the chemistry of l7a-hydroxy or 17a- acyloxy compounds. FMPT also activated the 16aJ7a-

epoxy-16b-methyl 17/3-carboxylic acids 14 to reaction with H2S. The resulting carbothioic acid 15 was unstable but treatment with bromochloromethane in situ gave the chloromethyl carbothioate 16. This could be rearranged with trifluoroacetic acid to the 16-methylene 17a-01 17 which on propionylation in the presence of acid gave a mixture of the l7a-ester log and the isomeric 11B-monoester (Scheme 5).

Activation of 14 with FMPT and reaction with NaSeH16 gave the corresponding carboselenoic acid, which could be alkylated i n si tu to the methyl or

christinehurd

Highlight


Scheme 5

Journal of Medicinal Chemistry, 1994, Vol. 37, NO. 22 3721

S M e 0 FMPT, U,N BCHzCI. b r r ~ [& Me

01) HIS

/ 0 15 16 14

CF,CO,H I COSCH,CI COSCH-CI

Scheme 6 COSeCH$

( i ) FMPT, Et3N {& CFJCO~H {&HZ I W ~ . ~ . OH

14 b Me - (ii) NaSeH

(iii) BrCHZCl or 198 X = H 18a XIH 19b XICI CHgl

18b XI CI

22a X = H 2la X - H 21b X = CI Nal

20a X = H 20b X = CI

23

chloromethyl carboselenoates 18a,b. Epoxide rearrangement (CF3COzH) then gave the corresponding 118- 17a-diols 19a,b. In this series, oxidation of 19a,b t o the 11-ketones 20a,b prior to 17a-propionylation followed by selective reduction of 11-oxo esters 21a,b t o the ll/3-alcohols 22a,b avoided unwanted llp-esterification (Scheme 6).

Reaction of the iodomethyl selenoester 22c with silver fluoride in acetonitrile gave the corresponding fluoromethyl selenoester 22d as a minor product (3%), but the major product (29%) was the 17P-acyl fluoride 23,

presumably formed by silver ion-assisted displacement of the selenide group by fluoride ion. While 23 was not analytically pure, the structure of the major component (83% by HPLC) was clearly revealed by its highly characteristic lH, 13C, and 19F NMR, infrared, and mass spectra. The same main fragment ions were observed in the chemical ionisation mass spectra of both 22d and 23, the initial losses being those of HSeCHzF and HF, respectively, from the major MH+ molecular ions. The MH+ peaks were confirmed by HRMS accurate mass measurements.


Table 3. Biological Activities of Halomethyl Androstane-17B-carbothioates and -17,9-carboselenoates

COSeCH& COSCH,X

" Y

10,11,12,13 22

mouse rat no. Z Y X R 16 human Va AITb HPAc AITb HPAc 13a H F F 10a F H C1 13b F H F 13c F H F 10b F F C1 13d F F F 13e F F F 13f F F F 1oc H H C1 13g H H F 10d F H C1 13h F H F 10e F H C1 12 F H Br

l l a F H I 13i F H F 10f F H C1 log F H C1 13j F H F 10h F F C1 13k F F F 22a F H H 22b F H C1

(fluocinolone acetonidelj

CZH5 €3 697 (230-2438) CZH5 €3 916 (471-1874) CzH5 H 1984 (1023-4013) CzH5 a-CH3 653 (311-1395) CzH5 a-CH3 124 (63-231) CH3 a-CH3 392 (159-959) CzH5 a-CH3 945 (551-1655) C3H7 a-CH3 299 (98-953) CzH5 B-CH3 295 (72-1148) CzH5 B-CH3 800 (231-2773) CH3 B-CH3 544 (154-2047) CH3 p-CH3 1388 (374-5370) CzH5 /3-CH3 1469 (858-2541) CzH5 B-CH3 254(G) CzH5 B-CH3 41(G) CzH5 P-CH3 1262 (597-2669) C3H7 p-CH3 143 (41-435) CzH5 =CHz 365 (198-665) CzH5 =CHz 1497 (575-4085) CzH5 =CHz 170(G) CzH5 =CHz 1048(G) CzH5 =CHz 187 (85-415) CzH5 =CHz 200 (103-388)

100

56 (36-86) 20 (14-29) 63 (39-105)

56 (37-86) 76d

113f 55 (34-87) 27 (17-44) 50 (35-71) 59 (33-103) 67 (47-95) 41 (27-64)

-

- -

89 (45-169)

42 (28-64) 41 (28-64)

108 (70-171) 197 (G) 27 (16-44) 40 (27-61)

-

100

>200 (GI 100 (G) 149 (103-219)

0.04 (0.01-0.09) 2.9 (1.4-5.2) 1.0 (0.5-2.1) 0.7 (G)

-

- - 14.7 (8.8-24.6) - 44 (GI

450h

2 (1-3) 44 (27-76)

- -

-

> 100 (GI > 100 (GI

2 (GI 100

-

97 (62-156)

39 (27-56)

29 (25-32)

859 55 (35-86) 29 (19-44) 44 (30-63)

40 (25-64) 36 (24-53)

-

-

24 (GI

- -

39 (30-52)

12 (9-17) 20 (14-26)

-

- -

19 (12-29) 16 (9-26)

100

1.5 (G) - '0.02 (G) 3 (GI 1.5 (0.8-2.2)

1.4 (0.5-3.2) 5.1 (1.8-11.8) 0.5 (G) 13.5 (G) 0.9 (0.5-1.7)

-

- - 13 (8-24)

0.8 (G) 2.5 (G)

-

- - <0.2

100 -

a Human vasoconstrictor activity relative to fluocinolone acetonide (100). Numbers in parentheses are 95% confidence level intervals or G represents a graphical estimate. Topical antiinflammatory activity relative to fluocinolone acetonide (100). Parentheses as for a. c Systemic corticosteroid activity after topical application relative to fluocinolone acetonide (100). Parentheses as for a. Mean of two results: 35 (21-58) and 117 (81-167). e Mean of three results: 103 (60-176), 84 (47-1581, and 76 (37-156). f Mean of two results: 134 (85-212) and 86 (46-150). g Mean of two results: 72 (57-90) and 97 (70-136). Mean of two results: 530 (G) and 370 (GI. Standard.

The lH NMR spectra of carbothioic acids were best measured in deuteriochloroform if solubility was sufficient, as in MezSO-ds impurity signals gradually appeared due to oxidation to the neutral disulfide (e.g., with 6i), a process complete in a longer time or at elevated temperature.

In each series of halomethyl l7a-acyloxy 17/3-carbothioates the ultraviolet extinction coefficients increased from ea. 16 000 to ca. 20 000, and the positions of the maxima moved slightly to higher wavelength (ca. 239 to 244 nm) on progressing from fluoromethyl through to iodomethyl thioesters. Gradations of optical rotation at 589 nm were also noted, increasing when a 16a- methyl substituent was present but decreasing when a 16P-methyl or no substituent was present. Presumably the contribution due to the carbothioate chromophore is considerably influenced by its conformation, and this in turn depends on the nature of ring D and its substituents.

Biological Results and Discussion Topical activity in humans was measured by the vas-

oconstriction assay according to the method of McKenzie and Atkins0n.l' Topical antiinflammatory activity was measured in rats and mice by modifications of the croton oil ear assay of Tonelli et a1.18 The undesired hypothalamic-pituitary-adrenal (HPA) function suppression was assessed in rats and mice by measuring reductions of levels of circulating corticosterone in response to ether

stress, using the procedure of Zenker and Bernstein.lg The results are shown in Table 3.

In the vasoconstriction assay, fluoromethyl carbothioates 13 were in general more active than their chloromethyl analogues 10 (13h, 10d; 13g, 1Oc; 13e, lob; 13b, loa; 13j, log; 13k, 10h; six pairs); however, for another pair (13i, 10e) the chloro analogue had the slightly greater potency. The bromomethyl and iodomethyl carbothioates 12 and l l a were less active than the chloromethyl analogue 10e. 17-Propionates 10e, 13e were better than the acetate (10d, 13d) and butyrate (lOf, 13f) analogues, although the potency of the propionate (13i) was marginally less than that of the acetate (13h). A similar pattern was found in the mouse and rat antiinflammatory activities. A different pattern was found for HPA suppression where the 16a- methyl compounds lob, 13d, and 13e showed little effect in both mouse and rat. The most active of these in the vasoconstriction and antiinflammatory tests was 13e (fluticasone propionate, FP), so this was chosen for more detailed examination. The carboselenoates 22a,b showed moderate activity but were not examined further as it was considered unlikely that drugs containing selenium would be acceptable.

The fluoromethyl and chloromethyl carbothioates 13i and 10e were, respectively, more than 3.5 and 1.5 times as active than their oxygen analogues1 in the vasoconstriction test, while the methyl carbothioate analogue of FP (13e) has been reportedll to show topical activity

Antiinflammatory Corticosteroid Analogues Journal of Medicinal Chemistry, 1994, Vol. 37, No. 22 3723

Sa-Fluoro-l1~,17a-dihydroxy-3-oxoandrosta-1,4-diene- 17B-carboxylic Acid (2b). A stirred suspension of Sa-fluoro- 11~-17a,21-trihydroxypregna-1,4-diene-3,20-dione (lby (10.0 g, 26.4 mmol) in tetrahydrofuran (55 mL) was stirred at 22 "C for 2 h with a solution of periodic acid (9.0 g, 39.5 mmol) in water (90 mL) and then poured into water (150 mL) and crushed ice (250 mL). Filtration gave 2b (9.42 g, 98%), a portion (0.100 g) of which was recrystallized from ethanol to give the analytical sample of 2b (0.060 g; 59%). 6a,Sa-Difluoro-11~,17a-dihydroxy-l6-methylene-3-

oxoandrosta-l,4-diene-l7~-carboxylic Acid (2c). 6a,9a- D~uoro-l1~,17a,2l-trihydroxy-l6-methylenepregna-l,4-diene- 3,20-dione (112)~~ (3.6 g) was oxidized as in the preceding experiment to give 2c (3.40 g, 98%). A portion (0.250 g) crystallized from aqueous methanol gave the analytical sample of 2c (0.162 g, 64%). 6a-Fluoro-11j3-hydroxyy-3-oxo-l7a-(propionyloxy)an-

drosta-l,4-diene-l7,5-carboxylic Acid (3a). A solution of 2a (4.49 g, 12.3 mmol) and Et3N (4.46 mL, 31.6 mmol) in CH2- Clz (160 mL) at -5 "C was treated dropwise with stirring with propionyl chloride (2.80 mL, 32 "01) in CHzClz (5 mL) during 5 min. After 20 min the reaction mixture was diluted with CHzClz, washed with aqueous NaHC03 and then HzO, dried, and evaporated to give the intermediate mixed anhydride as a solid (5.70 g). This was stirred in acetone (30 mL) with Etz- NH (4.6 mL, 44.3 mmol) for 30 min to give a clear yellow solution which was concentrated, diluted with water (150 mL), and washed with EtOAc. The aqueous phase was acidified to pH 2 with 2 N HCl(50 mL) and extracted with EtOAc. The extract was washed with water and evaporated to give a foam (5.82 g), a portion of which (0.304 g) was crystallized from EtOAc to give the analytical sample of 3a (0.144 g, 47%) as plates. Formation of 17P- [ (N,iV-dimethylcarbamoyl) thiol car-

bonyl Compounds 4b. General Method k The steroid 17P-carboxylic acid 3 (1 equiv) stirred in CHzClz (ca. 15 m u g of 3) was treated successively with Et3N (1 equiv) and NF- dimethylthiocarbamoyl chloride (2 equiv) under Nz at 20 "C, the reaction being monitored by TLC until no further con- sumption of starting material was observed (usually 6-30 h). The reaction mixture was diluted with EtOAc (ca. 50 m u g of 31, washed with 1N HC1, 5% NaHC03, and water, dried, and evaporated to give the crude anhydride 4b, purified by crystallization or by PLC then crystallization. 1784 [ (Na-Dimethylcarbamoyl)thiolcarbonyll-Sa-flu-

oro-11~-hydroxy-16a-methyl-l7a-(propionyloxy)androsta- 1,4-dien%-one (4ba). 9a-Fluoro-ll~-hydroxy-l6a-methyl-3- oxo-17a-(propionyloxy)androsta-1,4-diene-l7~-carboxylic acid (3bP was treated by general method A, but with the addition of NaI (1 equiv) to the reaction mixture, to give crystalline 4ba (63%) recrystallized twice from acetone to give the analytical sample: IR (CHBr3) 3610 (OH), 1740, 1710 (w) (COSCON, propionate), 1670, 1632, 1618 (A1s4-3-one) cm-l. 17a-Acetoxy- 178- [ [ (N,iV-dimethylcarbamoyl) thiol car-

bonyl]-Sa-fluoro-1 lfi-hydroxy-16,9-methylandrosta-1,4- dien-3-one (4bb). 17a-Acetoxy-9a-fluoro-ll/?-hydroxy-l6~- methylandrosta-l,4-diene-l7/3-carboxylic acid (3~)" treated by general method A gave crude anhydride (87%, ca. 90% pure by TLC), purified by PLC to give the analytical sample of 4bb as a foam: IR (CHBr3) 3590 (OH), 1735 (COSCON, propionate), 1665, 1630, 1612 (A1g4-3-one) cm-l. 17~-[[(N~-Dimethylcarbamoyl)thiolcarbonyll -Sa-flu-

oro-1 l~-hydro~y-16~-me~yl-l7a-(propionyloxy)andros~- 1,4-dien-3-one (4bc). 9a-Fluoro-ll~-hydroxy-l6~-methyl-3- oxo-l7a-(propionyloxy)androsta-1,4-diene-l7/?-carboxylic acid (3d),11 solvated with 0.75 mol of EtOAc (4.99 g, 9.96 mmmol) in pyridine (20 mL), was treated with dimethylthiocarbamoyl chloride (2.69 g, 21.8 mmol) for 19 h at room temperature. TLC indicated that only half of the starting material had been consumed. Isolations in general method A gave the crude anhydride (2.10 g, 40.5%), part of which was purified by PLC and crystallization from ether to give the analytical sample of 4bc: IR (CHBr3) 3670 (OH), 1740 (COSCON, propionate), 1670,1636, 1618 (A1s4-3-one) cm-'; lH NMR (90 MHz) 0.99 (13- CHa s) 1.07 (propionate CH3, t , J = 7 Hz), 1.29 (16P-CH3, d , J

in the vasoconstriction test only in the order of that of the standard, fluocinolone acetonide.

FP did cause HPA suppression if given as a suspension in saline by the subcutaneous route in rats [1.2 (0.8-1.713 and mice [0.7 (G)], compared with betamethasone (1.0); in rats the suppression was measured by weighing the adrenals. Given orally, however, it showed only weak adrenolytic activity in rats f0.13 (0.08-0.2l)l and depression of corticosterone levels in mice 10.01 (0.004-0.03)l compared with betamethasone (1.0). Weak glucocorticoid activity after oral administration is particularly of value in the treatment of airway conditions where a high proportion of the dose is swallowed.20

FP has been reported5 to be rapidly converted by liver homogenates (from mouse, rat, or dog) into the known carboxylic acid 3f.l' This acid has since been confirmed to be the principal metabolite in a study20 of the human pharmacology of FP, arising from orally administered drug by first-pass conversion in the liver, 3f showed negligible HPA suppression as measured by its adrenolytic activity in rats [<0.01 (GI1 and by depression of corticosterone levels in mice [<0.001 (G)] compared with betamethasone (1.0) by the subcutaneous route. For- mulations of FP have now received approval for the treatment of rhinitis (Flixonase21), asthma (in UK, Flixotide21), and steroid-responsive dermatoses (in USA, Cutivate21).

Experimental Section Melting points were determined on a Kofler block and are

uncorrected. Optical rotations were determined in dioxane at 20-22 "C; 'H NMR spectra were determined in MezSO-ds (unless stated otherwise) at 60, 90, or 100 MHz on Perkin- Elmer R24B, R32, or JEOL MHlOO spectrometers, respectively. The R32 instrument was also used at 84.68 MHz for 19F NMR. A JEOL FXlOO Fourier-transform spectrometer was used at 25.05 MHz for 13C NMR or at 100 MHz for 'H NMR. Chemical shifts are relative to Me3Si(CH~)sS03Na in MeZSO- d6 and SiMe4 in CHCl3-d for lH and 13C NMR and to CFCl3 for 19F NMR, as internal standards. IR spectra were recorded for Nujol mulls, or in CHBr3 where so indicated. Mass spectra were determined on a Finnigan MAT 4600 spectrometer using positive-ion chemical ionization with ammonia as the reagent gas. Accurate mass measurements were made on a Kratos Concept spectrometer by Mrs. V. Boote, Department of Chem- istry, University of Manchester. Organic reaction extracts were routinely dried over magnesium sulfate prior to removal of the solvent by rotary evaporation at ca. 20 mmHg at or below 50 "C. Products were dried in vacuo at up to 50 "C. Analytical TLC was conducted on Merck Kieselgel 60 F254 plates, developed with ch1oroform:acetone (e.g., 4:l) for neutral compounds or ch1oroform:acetone:acetic acid (e.g., 30:8:1) for acidic compounds. Preparative-layer chromatography (PLC) was performed in the same solvent systems on Merck Kieselgel 60 PFz54 + 366. Products were usually detected at 254 nm and eluted with ethyl acetate. Physical properties of intermediates are given in Table 1; those of biologically-assayed compounds are in Table 2. lH NMR (in MezSO-dj unless stated otherwise) and infrared spectra (in Nujol unless stated otherwise) were taken for all compounds, and where solvation is tabulated this was confirmed from the spectra. 6a-Fluoro-ll~,l7a-dihydroxy-3-oxoandrosta-1,4-diene-

l7P-carboxylic Acid (2a). A solution of Ga-fluoro-llP, 17a,- 2l-trihydroxypregna-1,4-diene-3,2O-dione (la)22 (4.99 g, 13.2 mmol) in tetrahydrofuran (50 mL) was stirred with a solution of periodic acid (10.0 g, 45 mmol) in water (24 mL) at 22 %oC for 50 min. The tetrahydrofuran was removed in uacuo to leave an aqueous suspension which was filtered, and the solid was washed with water and dried to give 2a (4.80 g, 100%). A portion (0.271 g) was crystallized from methanol to give the analytical sample of 2a (0.171 g, 63%).

3724 Journal of Medicinal Chemistry, 1994, Vol. 37, No. 22

= 6 Hz), 2.39 (propionate CH2, q, J = 7 Hz), 2.99, 3.04 (NMe2, singlets), 7.32 (1-H, d, J = 10 Hz).

In a second method, 9a-fluoro-l1/3-hydroxy-16~-methyl-3- oxo- 17a-(propionyloxy)androsta- 1 ,4-diene- 17/3-carbothioic acid, 6i, solvated with 0.75 mol of EtOAc (517 mg, 1.0 "01) and Et3N (0.49 mL, 3.5 mmol) in CH2Clz (10 mL), was treated with Nfl-dimethylcarbamoy1 chloride (0.37 mL, 4 mmol) and stirred under nitrogen at room temperature for 24 h. The mixture was diluted with EtOAc, washed successively with 1 N HCl, aqueous NaHC03, and HzO, dried, and evaporated to a foam (674 mg). PLC gave solvated 4bc as a solid (249 mg, 44%): [ a ] ~ +168" (c 1.05); MS mle 522 (MH+), 417 (MH+ - Me2- NCOSH), with mass, IR, and IH NMR spectra closely similar to those of the above. Anal. ( C Z ~ H ~ ~ F N O & O . ~ E ~ O A C ) C, H, N, S.

17p- [[ (N,iV-Dimethylcarbamo yl) thiol carbonyl] -9a-fluoro- 1 la-hydroxy- 16-methylene- 17a- (propiony1oxy)androsta-1,4-dien-3-one (4bd). Sa-Fluoro-11j3-hydroxy-16- methylene-3-oxo- 17a-(propionyloxy)androsta- 1 ,4-diene-1 78- carboxylic acid (3e)Z5 treated by general method A gave the crude anhydride (86%), a portion of which was purified by PLC and crystallization from acetone to give the analytical sample of 4bd (43%): IR (CHBr3) 3620 (OH), 1740, 1710 (COSCON, propionate), 1665, 1630, 1612 (A1r4-3-one) cm-l. 1 7 ~ - [ [ ~ ~ - ~ e ~ y l c a r b a m o y l ~ o x y l ~ ~ ~ l l - ~ - f l u o r o -

11/3-hydroxy-16/3-methyl- 17a-(propionyloxy)androsta- 1,4-dien-3-one (4ca). 9a-Fluoro-l1~-hydroxy-l6~-methyl-3- oxo-l7a-(propionyloxy)androsta-1,4-diene-l7~-carboxylic acid (3d),11 solvated with 0.75 mol of EtOAc (491 mg 0.98 mmol) in pyridine (2 mL), was treated with dimethylcarbamoyl chloride (208 mg, 1.94 mmol) at room temperature for 2 h. Isolation as in general method A gave the crude anhydride (489 mg), which was purified by two recrystallizations from acetone to give the analytical sample of 4ca (73%): IR (CHBr3) 3625 (OH), 1783 (COOCON), 1740, 740 (propionate), 1673, 1634, 1618 (A1z4-3-one) cm-'.

Formation of 17a-Hydroxy 17B-Carbothioic Acids 5. General Method B. A solution of the l7a-hydroxy 178- carboxylic acid 2 (1 equiv) in DMF (ca. 20 m u g of 2) was treated with 1,l'-carbonyldiimidazole (2 equiv), and the mixture was stirred under nitrogen at room temperature for ca. 4 h. Hydrogen sulfide was bubbled into the reaction for 15-30 min. After 0.5-4 h the reaction mixture was poured into 2 N HC1 and ice, and the precipitated acid 5 was collected by filtration.

Sa-Fluoro-1 1/3,17a-dihydroxy-3-oxoandrosta-1,4-diene- l7p-carbothioic Acid (Sa). The carboxylic acid 2b treated by general method B gave Sa (97%), mp 219-222 "C. A portion (0.12 g) crystallized from ethanol gave the analytical sample 5a (0.07 g, 57%). 9a-Fluoro-11~,17a-dihydroxy-16a-methy1-3-oxoandros-

ta-1,4-diene-17P-carbothioic Acid (Sb). 9a-Fluoro-llj3,17a- dihydroxy- 16a-methyl-3-oxoandrosta- 1,4-diene- 17a-carboxylic acid (2d)* by general method B gave the analytical sample of Sb (97%). 6a,Sa-Difluoro-l1/3,17a-dihydroxy- 16a-methyl-3-oxoan-

drosta-l,4-diene-l7~-carbothioic Acid (Sc). Ga,Sa-Dif- luoro-l1,5,1 7a-dihydroxy-l6a-methyl-3-oxoandrosta-l,4-diene- 178-carboxylic acid (2e)ll by general method B gave the analytical sample of Sc (92%): lH NMR (100 MHz) 0.85, (16a- CH3, d, J = 6 Hz), 0.97 (13-CH3, s), 1.50 (10-CH3, s), 4.20 (lla- H, m), 5.35 (llP-OH, m), 6.10 (4-H, m), 6.26 (2-H, dd, J = 10, 2 Hz), 7.26 (1-H, broad d, J = 10 Hz).

Sa-Fluoro- 11~,17a-~droxy-l6~-methyl-3-oxoandrosta- 1,4-diene-17/3-carbothioic Acid (Sd). 9a-Fluoro-l1~,17a- dihydroxy-16~-methyl-3-oxoandrosta-1,4-diene-l7~-carboxylic acid (2f)ll by general method B, except that the first stage was carried out at -5 "C for 18 h and the product was isolated by extraction with EtOAc and crystallized twice from EtOAc to give Sd (60%): lH NMR (90 MHz) 1.02 (13-CH3, s), 1.10 (16P-CH3, d, J = 6 Hz), 1.56 (10-CH3, s), 4.22 ( l la -H, m), 4.09 (4-H, m), 6.28 (2-H, dd, J = 10, 2 Hz), 7.37 (1-H, d, J = 10 Hz).

6a,9a-Difluoro- l lp , l7a-d ihydroxy- 16-methylene-3- oxoandrosta-1,4-diene-l7~-carbothioic Acid (Se). The carboxylic acid 2c by general method B gave 5e (89%). A

Phillipps et al.

portion (0.25 g) crystallized from ethyl acetate gave the analytical sample of Se (0.10 g, 36%).

Formation of 17a-Acyloxy 17B-Carbothioic Acids 6. General Method C. The crude 17j%[(N&-dimethylcarbamoyl)thio]carbonyl compound 4b prepared by general method A was refluxed under nitrogen in E t N (5-20 d g ) for 3.5-6 h. The cooled reaction mixture was poured into 6 N HCl (ca. 60 mug) and ice (ca. 60 mug). The product was extracted into EtOAc, and the extract was washed with water before back-extraction into 5% Na2C03 (ca. 120 mug). The aqueous layer was acidified with HC1 to pH 1 and the product extracted into EtOAc, washed with water, dried, and evaporated to give 6.

General Method D. The crude l7a-hydroxy 17p-carbothioic acid (S), prepared by general method B, with Et3N (ca. 3.5 equiv) in CHzCl2 (ca. 25 mug) at ca. 0 "C was treated dropwise with an acyl chloride (ca. 4.5 equiv) and then stirred for ca. 45 min. The solution was washed with 2 N NazC03, water, 2 N HCl, water, and brine, dried, and evaporated to leave a residue of the intermediate anhydride. This was dissolved in acetone (ca. 25 mug) and treated with Et2NH (ca. 10 equiv) for ca. 1 h and then poured into 2 N HC1 (ca. 40 mug) and ice (ca. 40 mug) to precipitate the carbothioic acid 5.

6a-Fluoro- 11/3-hydroxy-3-oxo- 17a-(propionyloxy)androsta-1,4-diene-l7/3-~arbothioic Acid (6a). The l7p-carboxylic acid 3a was converted by general methods A and then C and the product crystallized from a mixture of acetone and petrol (bp 60-80 "C) to give Sa (46%).

Sa-Fluoro- 1 l~-hydroxy-3-oxo-17a-(propionyloxy)androsta-1,4-diene-17/3-carbothioic Acid (6b). The l7p-carboxylic acid 2b was converted by general methods B then D (using propionyl chloride) to give crude 6b (93%), mp 118- 120 "C. A portion (0.35 g) crystallized from EtOAc gave the analytical sample of 6b (0.16 g, 44%).

Sa-Fluoro- 1 l j3 -h~ droxy- 16a-methyl-3-oxo- 17a-(propionyloxy)androsta-l,4-diene-l7~-carbothioic Acid (6c). 9a- Fluoro-l1/3-hydroxy-16a-methyl-3-oxo-l7a-(propionyloxy)androsta-l,4-diene-l7/3-carboxylic acid (3b)25 was converted by general methods A then C, and the product was crystallized twice from acetone to give 6c (29%). The 17a-hydroxy compound Sb also gave 6c (go%), mp 134-137 "C, by general method D, using propionyl chloride. 17a-Acetoxy-6a,9a-difluoro-l la-hydroxy- 16a-methyl-3-

oxoandrosta-l,4-diene-17/3-carbothioic Acid (6d). The 17a-hydroxy compound 5c was converted by general method D using acetyl chloride to give 6d (94%). Crystallization of a portion (0.40 g) from EtOAc gave the analytical sample of 6d (0.28 g, 64%).

6a,9a-Difluoro-l l,!?-hydroxy-16a-methy1-3-oxo-l7a-(propionyloxy)androsta-1,4-diene-17~-carbothioic Acid (6e). The 17a-hydroxy compound Sc was converted by general method D using propionyl chloride to give 6e (91%). Crystal- lization of a portion (0.40 g) from EtOAc gave the analytical sample of 6e (0.29 g, 54%): lH NMR (90 MHz, in CHC13-d) 0.99 (16a-CH3, d, J = 7 Hz), 1.15 (propionate CH3, t, J = 7 Hz), 1.16 (13-CH3, s), 1.54 (10-CH3, s), 2.40 (propionate CH2, q, J = 7 Hz), 4.46 (l la-H, broad m), 5.46 (6-H, broad dm, ca. J = 50 Hz), 6.40 (2-H, broad d, J = 10 Hz), 6.46 (4-H, m), 7.18 (1-H, d, J = 10 Hz). 6a,9a-Difluoro-ll~-hydroxy-16a-methyl- 3-oxo- 17a-(propionyloxy)androsta- 1,4-diene- 17~-carboXylic acid (3D11 also gave 6e (25%) using general methods A then C. 17a-(Butyryloxy)-6a,9a-difluoro-l la-hydroxy- 16a-meth-

yl-3-oxoandrosta-1,4-diene-l7~-carbothioic Acid (60. The 17a-hydroxy compound 5c was converted by general method D using butyryl chloride to give 6f (89%). Crystallization of a portion (0.40 g) from EtOAc gave the analytical sample of 6f (0.27 g, 60%).

1 1/3-€Iydroxy- 16,B-methyl-3-0~0- 1 7a-(propionyloxy)androsta-l,4-diene-l7~-carbothoic Acid (6g). 1 la-Hydroxy- 16/3-methyl-l7~-(propionyloxy)-3-oxoandrosta-l,4-diene-17~- carboxylic acid (3g)8 was converted by general methods A then C and crystallized from ethyl acetate to give 6g (24%).

17a-A~et0~y-9a-fluor0- 1 l~-hydroxy-16~-methyl-3-oxoandrosta-1,4-diene-l7~-carbothioic Acid (6h). Crude 4bb was converted by general method C to give 6h (51%). A

christinehurd

Highlight

christinehurd

Highlight

christinehurd

Highlight

christinehurd

Highlight

christinehurd

Highlight


(0.167 g, 1.99 mmol) in CHJ (0.3 mL, 4.8 mmol) and dimethylacetamide (2 mL) was stirred at room temperature for 3 h, diluted with EtOAc (150 mL), washed with 10% sodium thiosulfate (50 mL), water, 1 N HCl, water, 5% NaHC03, and water, dried, and evaporated in uacuo to give a foam (0.221 g) which was crystallized twice from methanol to give 9 (0.191 g, 67%): mp 235-237 "C; Amax (EtOH) 239 nm ( E 18 995) [lit." mp 223-224 "C; Amax (MeOH) 239 nm ( E 18 8901. Alterna- tively, 3d (0.681 g, 1.36 "01) and CDI (0.456 g, 2.8 mmol) were stirred under Nz at 22 "C in DMF (26 mL) for 19.5 h. A solution made from NaH (60% in oil, 0.285 g) in DMF (10 mL) saturated with MeSH was added, and after 5.5 h EtOAc (100 mL) was added and the solution was washed with 2 N- HC1, water, 2 N- Na2C03, HzO, and brine, dried, and evaporated in uucuo t o give 9 (0.306 g, 45%). Similarly, 7a (0.131 g) with MeSNa in DMF for 1 h gave crude 9 as a foam (0.103 g, 81%) and 7b (0.042 g) gave crude 9 (0.041 g, 104%). Formation of S-Chloromethyl 17P-Carbothioates 10.

General Method E. A solution of the carbothioic acid 6 (1 equiv) in dimethylacetamide (3.5-10 mug) was stirred with NaHC03 (2 equiv) and bromochloromethane (2-5 equiv) for 1-2 h, diluted with EtOAc (ea 200 mug), washed with 5% NaHC03 and water, dried, and evaporated in uucuo to give crude 10. S-Chloromethyl Sa-Fluoro-1 l~-hydroxy-3-oxo-l7a-

(propionyloxy)androsta-1,4-diene-17~-carbothioate (loa). General method E on 6a and two crystallizations from methanol gave 10a (52%). S-Chloromethyl 6a,9a-Difluoro-l lP-hydroxy-16a-meth-

yl-3-OXO- 17a-(propionyloxy)androsta- l,4-diene- 17p-carbothioate (lob). General method E on 6e and crystallization from EtOAc gave lob (67%): 'H NMR (90 MHz) 0.94 (16a- CH3, d, J = 7 Hz), 1.06 (13-CH3, s), 1.06 (propionate CH3, t, J = 7 Hz), 1.54 (10-CH3, s), 2.40 (propionate CHz, q, J = 7 Hz), 4.29 ( l la-H, broad m), 5.22 (CH2C1, s), ca. 5.65 (6-H, broad dm, J = 50 Hz), 6.18 (4-H, m), 6.36 (2-H, broad d, J = 10 Hz), 7.31 (1-H, broad d, J = 10 Hz). S-Chloromethyl 1 ID-Hydroxy- 168-methyl-3-oxo- 17a-

(propionyloxy)androsta-1,4-diene-l7~-carbothioate (10~). General method E on 6g and crystallization from EtOAc gave 1Oc (76%). S-Chloromethyl 17a-Acetoxy-9a-fluoro-1lj3-hydroxy-

16~-methyl-3-oxoandrosta-1,4-diene-17~-carbothioate (10d). General method E on 6h and two crystallizations from EtOAc gave 10d (76%). S-Chloromethyl Sa-Fluoro- 1 lp-hydroxy-16,t?-methy1-3-

oxo-17a-(propionyloxy)androsta-1,4-diene-l7~-carbothioate (10e). General method E on 6i and two crystallizations from methanol gave 10e (36%): 'H NMR (90 MHz) 0.98 (13- CH3, s), 1.06 (propionate CH3, t , J = 7 Hz), 1.37 (16P-CH3, d, J = 7 Hz), 1.53 (10-CH3, s), 2.38 (propionate CH2, q, J = 7 Hz), 4.28 (lla-H, broad m), 5.11 (CHZCl, s), 6.05 (4-H, m), 6.27 (2-H, broad d, J = 10 Hz), 7.29 (1-H, d, J = 10 Hz). A similar result was obtained with ICHzCl in place of BrCHZC1. S-C hloromethyl 17a-(Butyryloxy)-9a-fluoro- 1 ID-hy.

droxy-16,6-methyl-3-oxoandrosta-1,4-diene- 178-carbothioate (100. General method E on 6j and crystallization from EtOAc gave 10f (53%). S-Chloromethyl 9a-Fluoro-ll/3-hydroxy-16-methylene-

3-oxo- 17a-(propionyloxy)androsta- 1,4-diene- 17p-carbothioate (log). General method E on 6k and PLC followed by two crystallizations from EtOAc gave log (50%): 'H NMR (90 MHz) 1.01 (propionate CH3, t, J = 7 Hz), 1.01 (13-CH3, s), 1.54 (10-CH3, s), 2.32 (propionate CHZ, q, J = 7 Hz), 4.30 ( l la-H, m), 5.21 (CH2C1, s), 5.50-5.71 (l6=CHz and 11/3-OH, m), 6.08 (4-H, m), 6.29 (2-H, dd, J = 10, 2 Hz), 7.32 (1-H, d, J = 10 Hz). S-Chloromethyl Ga,Sa-Difluoro- 1 18-hydroxy- 16-meth-

ylene-3-oxo-17a-(propionyloxy)androsta-1,4-diene-17~- carbothioate (10h). General methods D then E on 5e, with PLC and crystallization from aqueous methanol, gave 10h (19.7%).

S-Chloromethyl Ga-Fluoro-1 l/?-hydroxy-3-oxo- 17a-(propionyloxy)androsta-1,4diene-17/3-carbothioate (1Oi). Gen- eral method E on 6a and crystallization from acetone and

portion (0.150 g) crystallized twice from EtOAc gave the analytical sample of 6h (0.127 g, 43%). 9a-Fluoro- 1 lj3-hydroxy-16j3-methyl-3-oxo-l7a-(propio-

nyloxy)androsta-l,4-diene-l7/3-carbothioic Acid (6i). Crude 4bc was converted by general method C and crystallization to give 6i (42%), mp 172-173 "C; a portion was recrystallized twice from acetone to give the analytical sample of 6i: 'H NMR (90 MHz) 1.00 (13-CH3, s), 1.07 (propionate CH3,

2.37 (propionate CH2, q, J = 7 Hz), 4.30 ( l la-H, broad m), 5.45 (llP-OH, m), 6.07 (4-H, m), 6.27 (2-H, d, J = 10 Hz), 7.32 (1-H, d, J = 10 Hz) with signals for the disulphide developing with time; lH NMR (90 MHz in CHC13-d) 1.07 (13-CH3, s), 1.17 (propionate CH3, t, J = 7 Hz), 1.37 (16P-CH3, d, J = 6 Hz), 1.57 (10-CH3, s), 2.38 (propionate CHZ, q, J = 7 Hz), 4.46 (lla- H, broad m), 6.17 (4-H, m), 6.37 (2-H, d, J = 10 Hz), 7.27 (1- H, d, J = 10 Hz).

The l7a-hydroxy compound 6d was also converted into 6i (53%), mp 174-179 "C from EtOAc, using general method D with propionyl chloride. In a third method, the carboxylic acid 3d11 (0.70 g, 1.40 mmol) and CDI (0.473 g, 2.9 mmol) in DMF (26 mL) were stirred at 22 "C under Nz for 19.5 h. A dark- blue solution, prepared by passing HzS through sodium hydride (60% in oil, 0.235 g) in DMF (10 mL), was added, and stirring was continued for 5.5 h. The mixture was diluted with EtOAc (100 mL), washed with 2N HC1, water, and brine, dried, and evaporated in uucuo t o give 6i (0.186 g, 26%). 17a-(Butyryloxy)-9a-fluoro-l lp-hydroxy-16/l-methy1-3-

oxoandrosta-l,4-diene-l7~-carbothioic Acid (6j). 17a- (Butyryloxy)-9a-fluoro-ll/3-hydroxy-16~-methyl-3-oxoandrosta- 1,4-diene-17P-carboxylic acid (3hI8 was converted by general methods A then C and crystallization from EtOAc to give 6j (34%). Sa-Fluoro-1 l~-hydroxy-16-methylene-3-oxo-l7a-(pro-

piony1oxy)androsta- lp-diene- 17,karbothioic Acid (6k). Crude 4bd was converted by general method C to give crude 6k (44%), purified by crystallization of a portion (0.20 g) from EtOAc to give the analytical sample of 6k (0.137 g, 30%). Sa-Fluoro-1 l~-hydroxy-16/3-methyl-3-oxoandrosta- 1,4-

diene-l7(R)-spiro-8-[2~-ethyl-2~-(imidazol-l-yl)- 1',3-di- oxolan-l'-one], Isomers 7a and 7b. A solution of 3d (1.337 g, 2.67 mmol) in DMF (40 mL) was stirred under Nz and treated with CDI (0.910 g, 5.6 mmol). After 5 h a t 22 "C the reaction mixture was poured into water (250 mL) and extracted with EtOAc. The extract was washed with water, 3% NazC03, water, and brine, dried, and evaporated in uacuo. PLC gave the more-polar isomer 7a (0.41 g, 32%), crystallized from EtOAc: IR 1810 (@-one), 1660, 1620, 1604 (A1,4-3-one) cm-l; 'H NMR (90 MHz) 0.85 (ethyl CH3, t, J = 8 Hz), 1.14 (13-CH3, s), 1.17 (l6P-CH3 d, J = 7 Hz), 1.47 (10-CH3, s), 2.31 (ethyl CH2, q, J = 8 Hz), 7.02,7.27,7.81 (imidazole protons); I3C NMR 8.6,34.8 (ethyl), 17.1 (ClS), 24.4 (C19), 71.4 (Cll) , 92.1 (C17), 102.1 (C9), 112.1 (CY), 117.7, 130.8, 135.3 (imidazole), 153.8 (Cl), 170.2 ((24'). The less-polar isomer 7b (0.218 g, 17%) crystallized from EtOAc: IR 1810 (@-one), 1667, 1629, 1604 (A1s4-3-one) cm-l; 'H NMR (90 MHz) 0.82 (ethyl CH3, t, J = 8 Hz), 0.85 (l6P-CH3, d, J = 7 Hz), 1.25 (13-CH3, s), 1.56 (10- CH3, s), 2.27 (ethyl CHZ, q, J = 8 Hz), 7.02, 7.37, 7.87 (imidazole protons); I3C NMR 8.6,34.0 (ethyl), 17.2 (C18), 24.5 (C19), 71.6 (Cll) , 91.1 (C17), 102.4 (C9), 112.0 (C2'), 118.0, 131.0, 136.0 (imidazole), 154.0 (Cl), 170.7 (C4'). In a repeat experiment on (0.455 g, 1.05 mmol) in which dimethylamine (1 mL) in DMF (9 mL) was added after 4 h, and stirring continued for 24 h, the isomers 7a and 7b were accompanied by a compound of intermediate polarity which crystallized from EtOAc-petroleum ether (bp 60-90 "C) to give 1-[(Sa-fluoro- 1 l~-hy~xy-l6~-methyl-3-oxo-l7a-propionylo~andros~- l,4-dien-l7~-yl)carbonyllimidazole (8, 0.039 g, 4%): IR 1742 (imidazole C=O and ester), 1669, 1625, 1605 (A1s4-3-one) cm-'; lH NMR (90 MHz) 0.94 (propionate CH3, t, J = 7 Hz),

2.43 (propionate CH2, q, J = 7 Hz), 7.08, 7.69,8.31 (imidazole protons). S-Methyl 9a-Fluoro-ll~-hydroxy-l6/3-methyl-3-oxo-

17a-(propionyloxy)androsta-1,4-diene-17~-carbothioate (9). A mixture of 6i (0.277 g, 0.61 mmol) and NaHC03

t, J = 7 Hz), 1.29 (16j3-CH3, d, J = 6 Hz), 1.54 (10-CH3, s),

1.08 (13-CH3, s), 1.35 (16P-CH3, d, J = 7 Hz), 1.54 (1O-CH3, s),


petroleum ether (bp 60-80 "C) then EtOAc and petroleum ether (bp 60-80 "C) gave 1Oi (81%). S-Chloromethyl Sa-Fluoro- 1 lp-hydroxy - 16a-methyl-3-

oxo-17a-(propionyloxy)androsta-l,4-diene- 17p-carbothioate (lOj). General method E on 6c and two crystallizations from acetone gave lOj (36%). S-Chloromethyl 17a-Acetoxy-6a,9a-difluoro-11~-hy.

droxy- 16a-methyl-3-oxoandrosta-1,4-diene-l7~-carbothioate (10k). General method E on 6d and crystallization from acetone gave 10k (74%). S-Chloromethyl 17a-(Butyryloxy)-6a,9a-difluoro-11~-

hydroxy-16a-methy1-3-oxoandrosta- 1,4-diene- 17p-carbothioate (101). General method E on 6f and two crystallizations from acetone gave 101 60%). Formation of S-Iodomethyl17j3-Carbothioates 11. Gen-

eral Method F. The S-chloromethyl carbothioate 10 (1 equiv) and NaI (ca. 4 g/g, ca. 12 equiv) were refluxed in acetone (10- 30 mug) for 3-7 h. EtOAc (150 mug) was added, and the solution was washed with water, 10% sodium thiosulfate or 5% sodium metabisulfate, 5% NaHC03, and water, dried, and evaporated in uacuo t o give crude 11. S-Iodomethyl 9a-Fluoro-ll~-hydroxy-l6~-methyl-3-

oxo-17a-(propionyloxy)androsta-1,4-diene-l7~-carbothioate (Ila). General method F on 10e and PLC and then two crystallizations from acetone gave lla (41%): 'H NMR (90 MHz in CHCl3-d) 4.26 and 4.52 (CHzI, dd, J = 11 Hz). S-Iodomethyl 6a-Fluoro-ll~-hydroxy-3-oxo-l7a-(pro-

pionlyoxy)androsta-l,4diene-l7~-carbothioate (llb). Gen- eral method F on 1Oi and two crystallizations from acetone and petroleum ether (bp 60-80 "C) gave llb (76%). S-Iodomethyl 9a-Fluoro- 1 l~-hydroxy-3-0~0- 17p4pro-

pionyloxy)androata-l,4diene-l7~~~bothioate (1 IC). Gen- eral method F on 10d and crystallization from methanol gave llc (81%). S-Iodomethyl17a-Acetoxy-6cq9a-duoro-ll~-hydroxy-

16a-methyl-3-oxoandrosta-1,4-diene-l7~-carbothioate (Ild). General method F on 10k and crystallization from EtOAc gave lld (82%).

S-Iodomethyl6a,9a-Difluoro- 1 lp-hydroxy - 16a-methyl- 3-oxo-l7a-(propionyloxy)an~osta-l,4-diene-l7~-carbothioate (lle). General method F on 10b and crystallization from EtOAc gave lle (85%): 'H NMR (90 MHz) 4.63 (CHzI, SI. S-Iodomethyl17a-(Butyryloxy)-6a,9a-duoro-1 lp-hy-

droxy- 16a-methyl-3-oxoandrosta- l,a-diene- 17p-carbothioate (110. General method F on 101 and crystallization from EtOAc gave llf (85%). S-Iodomethyll7a-Acetoxy-9a-fluoro- 1 113-hydroxy- 16p-

methyl-3-oxoandrosta-l,4-diene-l7~-carbothioate (llg). General method F on 10d then PLC and two crystallizations from EtOAc gave llg (71%). S-Iodomethyl Sa-Fluoro-1 I/?-hydroxy- 16-methylene-3-

oxo- 17a- (propiony1oxy)androsta- l,a-diene- 17p-carbothioate (llh). General method F on log then PLC and two crystallizations from acetone gave llh (74%). S-Bromomethyl Sa-Fluoro- 1 lp-hydroxy- 16p-methyl-3-

oxo- 17a- (propiony1oxy)androsta- 1,4-diene- 17P-carbothioate (12). A solution of lla (0.660 g, 1.12 mmol) in acetone (20 mL) was stirred with LiBr (0.972 g, 11.2 mmol) at room temperature for 5 d. The reaction mixture was diluted with EtOAc (150 mL), washed with 10% sodium thiosulfate, water, and brine, dried, and evaporated in uacuo to give a foam (0.624 g). This crystallized from acetone and petroleum ether (bp 40- 60 "C) to give 12 (0.499 g, 82%): 'H NMR (90 MHz in CHC13- d ) 4.51 and 4.89 (CHZBr, dd, J = 11 Hz). Formation of S-Fluoromethyl 17P-Carbothioates 13.

General Method G. The S-iodomethyl carbothioate (11, 1 equiv) and AgF (3-10 equiv) were stirred in the dark at room temperature in CHsCN (12-75 mug) for 1-72 h. The highest reaction rates were achieved with finely pulverized AgF, obtained in granular form (from Ventron). The reaction mixture was diluted with EtOAc (ca. 150 mug) and filtered through kieselguhr. The filtrate was washed with water, dried, and evaporated in vacuo to give crude 13. S-Fluoromethyl6a-Fluoro-1 lj3-hydmxy-3-0xo-l7a-(pro-

pionyloxy)androeta-l,4-diene-l7p-car~t~oa~ (13a). Gen-

Phillipps et al.

era1 method G on llb and two crystallizations from acetone and petroleum ether (bp 60-80 "C) gave 13a (62%). S-Fluoromethyl Sa-Fluoro-1 lp-hydroxy-3-0xo-l7a-(pro-

pionyloxy)androstra-l,4-diene-17~-carbothioate (13b). General method G on llc and two crystallizations from MeOH gave 13b (58%). S-Fluoromethyl Sa-Fluoro- 1 lp-hydrosy-16a-methy1-3-

oxo-l7a-(propionyloxy)androsta-l,4-diene-17~-carbothioate (13c). General methods F then G on lOj, without purification of the intermediate S-iodomethyl carbothioate, PLC, and two crystallizations from acetone gave 13c (43%). S-Fluoromethyl 17a-Acetoxy-6a,9a-difluoro-11~-hy-

droxy-l6a-methyl-3-oxoandrosta-1,4-diene-17~-carbothioate (13d). General method G on lld and recrystallization from EtOAc gave 13d (70%). S-Fluoromethyl6a,9a-Difluoro-ll~-hydroxy-16a-meth-

yl-3-oxo- 17a-(propionyloxy)androsta- 1,4-&ene- 17p-carbothioate (13e). Small pieces ofAgF (2.46 g, 19.4 "01) were stirred in CH3CN (320 mL, dried over molecular sieves) for 17 h. The S-iodomethyl ester lle (5.9 g, 9.7 mmol) was added, and the reaction mixture was stirred for 2 h. Finely powdered AgF (2.46 g, 19.4 mmol) was added, and after being stirred for a further 2 h the reaction mixture was concentrated to ca. 50 mL in vacuo and diluted with EtOAc (800 mL). After filtration through kieselguhr, The filtrate was washed with 2 N HC1 (400 mL), water (3 x 400 mL), and brine (400 mL), dried over MgS04, and evaporated in vacuo to low volume when crystallization occurred. The white crystals (4.2 g, 87%) were collected and recrystallized from acetone to give 13e (3.6 g, 74%): IR 3350 (OH), 1750 (propionate), 1708 (carbothioate), 1668, 1622, 1612 (A1s4-3-one) cm-l; 'H NMR (90 MHz) 0.96 (16a-CH3, d, J = 7 Hz), 1.06 (propionate CH3, t, J = 7 Hz), 1.07 (13-CH3, s), 1.56 (10-CH3, s), 2.33 (propionate CHz, q, J = 7 Hz), 4.32 (lla-H, broad m), 5.63 (11-OH, broad d, J = 4 Hz), ca. 5.75 (6-H, broad dm, J = 50 Hz), 6.00 (SCHzF, d, J = 51 Hz), 6.19 (4-H, m), 6.37 (2-H, dd, J = 10,2 Hz), 7.33 (1-H, d, J = 10 Hz).

The fluoromethyl ester 136 was also prepared from the carbothioic acid 6e (0.5 g, 1.07 mmol) in DMF (2.05 mL) by first stirring it at -5 "C under nitrogen for 5 min with &CO3 (0.118 g, 1.18 mmol). Cold (-60 "C) bromofluoromethane (0.138 g, 1.22 mmol) was added, and the mixture was stirred a t 0 "C to -5 "C for 1 h and then diluted with EtOAc (6 mL). The mixture was washed with 5% NazC03, and the aqueous layer was extracted with EtOAc (6 mL). The combined organic extracts were washed with HzO (4 mL) and concentrated in uacuo to ca. 1.5 mL. After the suspension was cooled to 0 "C and stirred for 1 h, the solid was collected by filtration and dried at 40 "C in vacuo for 16 h to give 13e (0.370 g, 69.3%), with a second crop (0.050 g, 9.4%) by concentration of the mother liquors.

The fluoromethyl ester 13e was also prepared from the carbothioic acid 6e (47.9 g, 102 mmol) in DMF (190 mL) by first stirring it with KHC03 (11.26 g, 112 mmol) at room temperature for 10 min under Nz. Fluoroiodomethane (17.2 g, 107 mmol) was added over 3-4 min with cooling to keep the temperature at 22-25 "C. Isolation after 0.25 h as in the preceding experiment gave 13e (67.1%), mp 184-185 "C, with IR and 'H NMR spectra resembling those detailed above. The latter showed solvation with DMF (ca. 0.05 mol). S-Fluoromethyl 17a-(Butyryloxy)-6a,9a-difluoro-1 lp-

hydroxy-16a-methyl-3-oxoandrosta-l,4-diene-l7~-carbothioate (130. General method G on llf and crystallization from EtOAc gave 13f (76%). S-Fluoromethyl 1l~-Hydroxy-l6~-methyl-3-oxo-l7a-

(propionyloxy)androsta-l,4--dien~l7~-c~~~oate (13g). General methods F then G on lOc, without purification of the intermediate S-iodomethyl carbothioate, and crystallization from EtOAc gave 13g (23%). S-Fluoromethyl 17a-Acetoxy-9a-fluoro-1l~-hydroxy-

16~-methyl-3-oxoandrosta-1,4-diene-l7~-carbothioate (13h). General method G on llg and two crystallizations from acetone gave 13h (35%). S-Fluoromethyl Sa-Fluoro-1 lp-hydroxy-16p-methy1-3-

oxo- 17a-(propionyloxy)androsta- 1,4-diene- 17b-carboth-

christinehurd

Highlight


ioate (13i). General method G on lla with PLC and crystallization from MeOH then MeOH and Et20 gave 13i (33%).

The fluoromethyl ester 13i was also prepared from the chloromethyl ester 10e (0.377 g, 0.760 mmol) with AgF (0.669 g, 5.3 mmol) in CHsCN (6 mL) in the dark for 38 days. Isolation as in general method G gave 13i (0.070 g, 19.2%): lH NMR (90 MHz) 0.94 (13-CH3, s), 1.06 (propionate CH3, t, J

(propionate CHz, q, J = 7 Hz), 4.27 ( l la-H, m), 5.84 (CHzF, d, J = 51 Hz), 6.04 (4-H, m), 6.24 (2-H, dd, J = 10, 2 Hz), 7.29

= 51 Hz), -162 (9a-F, broad dd). S-Fluoromethyl Sa-Fluoro-1 lp-hydroxy-16-methylene-

3-oxo- 17a-(propionyloxy)androsta- 1,4-diene- 17B-carbothioate (13j). General method G on llh with PLC and two crystallizations from acetone gave 13j (22%). S-Fluoromethyl Ga,Sa-Difluoro-l lp-hydroxy-16-meth-

ylene-3-oxo-17a-(propionyloxy)androsta-l,4-diene- 178- carbothhioate (13k). General methods F then G on lOh, without purification of the intermediate S-iodomethyl carbothioate, PLC, and crystallization from aqueous acetone gave 13k (79%): lH NMR 0.98 (13-CH3, s), 1.03 (propionate CH3, t, J = 7 Hz), 1.55 (10-CH3, s), 2.34 (propionate CH2, q, J = 7 Hz), 4.32 (Ha-H, m), 5.63 (l6=CHz, broad m), ca. 5.7 (6-H, broad dm, J = 50 Hz), 5.95 (SCHzF, d, J = 51 Hz), 6.18 (4-H, m), 6.35 (2-H, dd, J = 10, 2 Hz), 7.30 (1-H, broad d, J = 10 Hz). S-Chloromethyl16a,17a-Epoxy-9a-fluoro-1 lj3-hydroxy-

16,f3-methyl-3-oxoandrosta-l,4-diene-17~-carbothioate (16). A suspension of 16a,17a-epoxy-9a-fluoro-l1~-hydroxy-16/3- methyl-3-oxoandrosta- 1,4-diene- 17p-carboxylic acid ( 14Y6 (0.753 g, 2.0 mmol) and FMPT (0.680 g, 2.4 mmol) in CHzClz (7 mL) was treated dropwise at 0 "C with Et3N (1.39 mL, 10 mmol) and then stirred at 0 "C for 1 h. HzS was passed through the mixture for 15 min, and the resultant solution was stirred at 0 "C for 1 h. After addition of BrCHzCl (0.26 mL, 4 mmol) the mixture was stirred for 1.5 h at room temperature, diluted with EtOAc (250 mL), washed with 2 N HC1, 5% NaHC03, and water, dried, and evaporated in vacuo to give a yellow solid (0.818 g). This on PLC and then crystallization from acetone gave 16 (51%). S-Chloromethyl Sa-Fluoro-1 lp,l7a-dihydroxy-16-meth-

ylene-3-oxoandrosta-1,4-diene-l7~-carbothioate (17). A solution of 16 (0.400 g, 0.91 mmol) in CF~COZH (16 mL) was stirred for 5.5 h at room temperature, evaporated to near dryness in vacuo, and dissolved in EtOAc (100 mL). The solution was washed with 5% NaHC03 and water, dried, and evaporated in uacuo to give a foam (0.466 g), part of which on PLC and two crystallizations from acetone gave 17 (70%).

The llpJ7a-diol 17 (0.227 g, 0.52 mmol) in propionic acid (2.2 mL) and (CF3CO)zO (0.7 mL) was treated with a dry solution ofp-toluenesulfonic acid in CHCl3 (0.044 mL, containing ca. 0.08 g/mL) and then stirred at room temperature for 6 h and at 3 "C for 16.5 h. The reaction mixture was diluted with EtOAc, washed with 5% NaHC03, H20, and brine, dried, and evaporated in vacuo to give a mixture of lip- and 17a- monopropionates. Purification by PLC gave sample of log (48%) containing ca. 7% of the 1lp-monopropionate. Se-Methyl 16a,17a-Epoxy-9a-fluoro-11/3-hydroxy-16,f3-

methyl-3-oxoandrosta-l,4diene-l7/3-carboselenoate (Ma). The 17~-carboxylic acid 14 (0.376 g, 1.0 mmol) and FMPT (0.340 g, 1.2 mmol) were stirred under NZ at 0 "C, and Et3N (0.70 mL, 5.0 mmol) was added dropwise. After 70 min a solution of NaSeH [prepared under Nz by the addition of EtOH (3 mL) to a mixture of NaBH4 (0.063 g, 1.66 mmol) and powered Se (0.118 g, 1.5 mmol) at 0 "C followed by stirring for 20 min] was added, and the brown solution was stirred at 0 "C for 1.25 h. Me1 (0.12 mL, 2.0 mmol) was added under Nz, and the yellow solution was stirred for 3.75 h at room temperature. The reaction mixture was diluted with EtOAc (200 mL), washed with 2 N HC1, 5% NaHC03, and water, dried, and evaporated in uacuo to give a solid (0.390 g). PLC and two crystallizations from acetone gave 18a (0.174 g, 38.4%): IR (in CHBr3) 1680 (carboselenoate) cm-l; lH NMR

= 7 Hz), 1.35 (16p-CH3, d, J = 7 Hz), 1.52 (10-CH3, s), 2.38

(1-H, d, J = 10 Hz); "F NMR (84.68 MHz) -187.5 (CH2F, t, J

Journal of Medicinal Chemistry, 1994, Vol. 37, No. 22 3727

(90 MHz) 1.38 (13-CH3, s), 1.54 and 1.58 (16P-CH3 and 10- CH3, singlets), 2.22 (SeCH3, s). Se-Chloromethyll6aJ7a-Epoxy-9a-fluoro-1 l/3-hydroxy-

16/3-methyl-3-oxoandrosta-l,4-diene-17/3-carboselenoate (18b). The carboxylic acid 14 (0.376 g, 1.0 mmol) reacted as in the preceding experiment, using FMPT-NEt3, NaSeH [from Se (2.0 mmol)] and BrCHzCl (0.130 mL, 2.0 mmol) in place of MeI, to give, after crystallization from acetone, 18b (29.5%): IR 1710 (carboselenoate) cm-l; 'H NMR (90 MHz) 1.41 (13-CH3, s), 1.54 and 1.58 (16p-CH3 and 10-CH3, singlets), 5.06 (SeCH2C1, s). Se-Methyl9a-Fluoro-1 l/3,17a-dihydroxy-l&methylene-

3-oxoandrosta-1,4-diene-l7/3-carboselenoate (19a). The epoxide 18a (0.574 g, 1.27 mmol) in CF~COZH (25 mL) was stirred at room temperature for 2.5 h and then diluted with 5% NaHC03 (550 mL). The product was extracted into EtOAc, and the extract was washed with water, dried, and evaporated in vacuo to give a yellow foam (0.561 g). PLC and two crystallizations from acetone gave 19a (0.326 g, 56.8%): IR 1700 (carboselenoate) cm-l; 'H NMR (90 MHz) 0.91 (13-CH3, s), 1.53 (10-CH3, s), 2.06 (SeCH3, SI, 4.97 and 5.22 (l6=CHz, broad singlets), 6.38 (17a-OH, s). Se-Chloromethyl ~-Fluoro-ll/3,17a-dihydroxy-16-meth-

ylene-3-oxoandrosta-1,4-diene-17/3-carboselenoate (19b). The epoxide 18b (1.332 g), in a manner similar to the preceding experiment, gave crystalline 19b (0.276 g, 20.7%): IR 1720 (carboselenoate) cm-'; IH NMR (90 MHz) 0.96 (13-CH3, s), 1.53 (10-CH3, s) 5.01 (SeCH2C1, s), 5.00 and 5.25 (l6=CHz, broad singlets), 6.71 (17a-OH, 8). Se-Methyl Sa-Fluoro- 17a-hydroxy- 16-methylene-3,ll-

dioxoandrosta-1,4-diene-l7/3-carboselenoate (20a). Py- ridinium dichromatez7 (0.130 g, 0.345 mmol) was added to a solution of 19a (0.125 g, 0.246 mmol) in DMF (1.4 mL), and the mixture was stirred at 0 "C for 6 h. More pyridinium dichromate (0.130 g, 0.345 mmol) was added, and stirring continued for 24 h at 3-4 "C. The mixture was diluted with water (20 mL), and the product was extracted into EtOAc. The extract was washed with water, dried, and evaporated in vacuo to a yellow gum (0.116 g) which, on PLC and two crystallizations from acetone, gave 20a (0.050 g, 45.0%): IR 3390 (OH), 1718 (11-one), 1689 (carboselenoate) cm-l; 'H NMR (90 MHz) 0.66 (13-CH3, s), 1.53 (10-CH3, s), 2.09 (SeCH3, s), 5.04 and 5.29 (l6=CH2, broad singlets), 6.92 (17a-OH, 8).

Se-Chloromethyl9a-Fluoro-17a-hydroxyy-16methylene- 3,11-dioxoandrosta-l,4-diene-l7/3-carboselenoate (20b). Pyridinium dichromate (7.825 g, 20.8 mmol) was added to a stirred solution of 19b (4.155 g, 8.32 mmol) in DMF (46 mL), and after 3.25 h the product was isolated as in the preceding experiment to give crude 20b (3.00 g, 74.2%). Two crystallizations of a portion (0.150 g) from acetone gave the analytical sample of 20b (0.056 g): IR 3350 (broad, OH), 1720 (11-one), 1709 (carboselenoate); 'H NMR (90 MHz) 0.83 (13-CH3, s), 1.54 (10-CH3, s), 5.06 (SeCHZC1, s), 5.06 and 5.34 (l6=CHz, broad singlets), 7.26 (17a-OH, 9). &-Methyl Sa-Fluoro-16-methylene-3,l l-dioxo-l7a-(pro-

pionyloxy)androsta-l,4-diene-l7/3-carboselenoate (21a). A solution of2Oa (1.451 g, 3.214 mmol) in propionic acid (14.5 mL) and (CF3CO)zO (5.8 mL) was stirred and treated with a solution ofp-toluenesulfonic acid (0.016 g) in CHC13 (0.20 mL) for 38 h at room temperature. The mixture was poured into 5% NaHC03 (400 mL), and the product was extracted into EtOAc (300 mL). The extract was washed with water, dried, evaporated in vacuo, and purified by PLC to give 21a (1.095 g, 67%). A portion of (0.250 g) recrystallized twice from acetone gave the analytical sample of 21a (0.198 g): IR 1740 (propionate), 1722 (11-one), 1678 (carboselenoate) cm-'; 'H NMR (90 MHz) 0.71 (13-CH3, s), 1.05 (propionate CH3, t, J = 7 Hz), 1.55 (10-CH3, s), 2.23 (SeCH3, SI, 2.39 (propionate CHZ, q, J = 7 Hz), 5.64 and 5.70 (l6=CHz, broad singlets). Se-Chloromethyl 9a-Fluor0-16-methylene-3,ll-dioxo-

17a-(propionyloxy)androsta-1,4-diene-l7/3-carboselenoate (21b). Propionylation of 20b (2.850 g, 5.87 mmol) as in the preceding experiment but for 6 days gave 21b (2.380 g, 75%). A portion (0.150 g) crystallized twice from acetone gave the analytical sample of 21b (0.070 g): IR 1744 (propionate), 1718 (11-one), 1688 (carboselenoate) cm-l; IH NMR (90 MHz)


0.77 (13-CH3, s), 1.06 (propionate CH3, t, J = 7 Hz), 1.55 (10- CH3, s), 2.42 (propionate CH2, q, J = 7 Hz), 5.21 (SeCH2C1, SI, 4.30 (l6=CHz, broad).

Se-Methyl Sa-Nuoro-1 ~-hydroxy-l6-methylene-3-0~0- 17a-(propionyloxy)androsta- 1,4-diene-l7/3-carboselenoate (22a). A suspension of 21a (0.841 g, 1.66 mmol) and NaBH4 (0.069 g, 1.82 mmol) in EtOH (12 mL) was stirred at room temperature for 1.3 h, treated with acetone (3.5 mL), and concentrated to near dryness in uacuo. EtOAc (100 mL) was added, and the solution was washed with 1N HC1 and water, dried, and evaporated in uacuo to give a white foam. Two crystallizations from acetone gave 22a (0.657 g, 77.7%): IR 3340 (broad, OH), 1744, 1735 (propionate), 1718 (carboselenoate) cm-l; 'H NMR (90 MHz) 0.96 (13-CH3, s), 1.02 (propionate CH3, t, J = 7 Hz), 1.54 (10-CH3, s), 2.20 (SeCH3, s), 2.32 (propionate CH2, q, J = 7 Hz), 5.61 and 5.66 (l6=CHz, broad singlets).

Se-Chloromethyl Sa-Fluoro-1 l@-hydroxy- lsmethylene- 3-oxo-17a-(propionyloxy)androsta- 1,4-diene- 17/3-carboselenoate (22b). The 11-ketone 21b (2.656 g, 4.90 mmol) was reduced as in the preceding experiment to give 22b as a white foam (1.802 g, 67.7%). A portion (0.500 g) was recrystallized twice from EtOAc to give the analytical sample of 22b (0.297 g): IR 3310 (broad, OH), 1740 (propionate), 1724 (carboselenoate) cm-'; IH NMR (90 MHz) 1.01 (13-CH3, s), 1.01 (propionate CH3, t, J = 8 Hz), 1.55 (10-CH3, s), 2.33 (propionate CH2, q, J = 8 Hz), 4.29 ( l la-H, m) 5.19 (SeCHZCl, s), 5.60 (118- OH, m), 5.62 (l6=CHz, broad).

Se-Iodomethyl Sa-Fluoro-1 1/3-hydroxy-16-methylene- ~-oxo-l7a-(propionyloxy)~drosta-1,4-diene-l7/3-~arbose- lenoate (22c). A mixture of the Se-chloromethyl selenoester 22b (1.291 g, 2.37 mmol) and NaI (4.447 g, 29.7 mmol) in acetone (40 mL) was heated under reflux for 24 h. The mixture was cooled, diluted with EtOAc (200 mL), washed with water (60 mL), 10% sodium thiosulfate (2 x 60 mL), 5% NaHC03 (60 mL), and water (2 x 60 mL), dried, and evaporated to a foam (1.392 g). PLC gave colorless crystals (1.225 g, 81%), a portion (0.225 g) of which was recrystallized twice from acetone to give 22c (0.201 g, 72%): IR 1744 (propionate), 1728 (carboselenoate) cm-l; 'H NMR (90 MHz) 1.00 (propionate CH3, t, J = 8 Hz), 1.02 (13-CH3, s), 1.53 (10-CH3, s), 1.00 (propionate CH3, t, J = 8 Hz), 1.02 (13-CH3 s), 1.53 (10- CH3, s), 2.31 (propionate CH2, q, J = 8 Hz), 4.43 (SeCH21, s), 5.53-5.71 (l6=CHz, 11-OH, broad).

Reaction of 22c with Silver Fluoride. 22c (1.0 g) and AgF (2.0 g, 10 equiv) in CH3CN (16 mL) were stirred at ambient temperature in the dark for 24 h. The reaction mixture was diluted with EtOAc (300 mL), solid material was removed by filtration, and the filtrate was washed with water, dried, and concentrated to a foam (0.690 g). Silica gel chromatography eluting with CHC13 gave a white solid (0.027 g, 3%, mechanical loss) which was recrystallized from EtOAc- petroleum ether (bp 40-60 "C) to give colorless crystals (0.008 g) of Se-fluoromethyl Sa-fluoro-ll/3-hydroxy-l6-methyl- ened-oxo- 17a- (propiony1oxy)androsta- 1 p-diene- 17B-c~~- boselenoate (22d): IR 3600-3000 (OH), 1738, 1708, 1695, (shoulder) cm-'; IH NMR (100 MHz) 6 0.96 (13-CH3, SI, 1.00 (propionate CH3, t, J = 7 Hz), 1.51 (10-CH3, s) 2.32 (propionate CH2, q, J = 7 Hz), 5.5-5.6 (l6=CH2, 11-OH, broad), 6.22 (SeCHzF, d, J = 50 Hz); MS mle (re1 intensity) Se cluster MH+ 531 (23.0), 529 (loo), 527 (52), 526 (19.4), 525 (19.9), 523 (3.0), fragments 415 (14.7), MH - SeCHzF), 359 (12.41, 313 (35.11, 293 (24.2). The MH+ ion-cluster had relative peak intensities close to the calculated values for the seven Se isotopes, HRMS mle calcd for C25H31F~O5*~Se (MH+) 529.1304, found 529.1269. Further elution gave a white foam (0.543 g, 80%), which was recrystallized twice from acetone to give colorless crystals (0.196 g, 29%) of 9a-fluoro-ll~-hydroxy-l6-methylene-3-oxo- 17a-(propionyloxy)androsta-1,4-diene-l7~-carbonyl fluoride 23: 83% pure by HPLC; mp 193-195 "C; [ a h -82" (c 0.93); A,, 239 nm (ElcmlB 346); IR (CHBr3) 3605 (OH), 1842 (COF), 1738 and 1262 (propionate), 1668, 1630, and 1612 (A1p4-3-one) cm-l; 'H NMR (90 MHz) 1.01 (propionate, t, J = 8 Hz), 1.10 (13-CH3, s), 1.54 (10-CH3, s), 2.38 (propionate, q, J = 8 Hz), 6.55 and 6.65 (l6=CHz, 11-OH, broad); I9F NMR (84.68 MHz) +29 (COF, s), -161.7 (9a-F, m); I3C NMR (25.05 MHz) 174.3

Phillipps et al.

(8, propionate C=O), 159.3 (d, 'JCF = 365 Hz, COF), 145.4 (6, C16), 120.9 (s, l6=CH2; off-resonance t), 87.8 (d, 'JCF = 50 Hz, C17), 10.1 (s, propionate CH3, off-resonance 9); MS mle (re1 intensity) 435 (98.7, MH+), 415 (100, MH+ - HF), 359 (35.7), 313 (16.7), 293 (15.8); HRMS mle calcd for C24HzsFz05 (MH+) 435.1983, found 435.1982. Anal. (Cz4H28'205) H, F; C calcd, 66.3; found, 65.3.

Bis[Sa-fluoro-1 l&hydroxy-16/3-methyl-3-oxo-l7a-(pro- piony1oxy)androsta- 1,4-dien-l7/3-yl)carbonyll Disulfide. The carbothioic acid 6i (0.104 g, 0.230 mmol) in DMSO (1 mL) was stirred and heated at 85 "C for 1.5 h. The solution was diluted with EtOAc (100 mL), washed with 5% NaHC03 and water, dried, and evaporated in uacuo to a solid (0.108 g). PLC and two crystallizations from acetone gave the title disulfide

(propionate CH3, t, J = 7 Hz), 1.32 (16P-CH3, d, J = 6 Hz), 1.52 (10-CH3, s), 2.42 (propionate CH2, q, J = 7 Hz), 4.33 (lla- H, m), 5.72 (llP-OH, m), 6.08 (4-H, s), 6.29 (2-H, d, J = 10 Hz), 7.36 (1-H, d, J = 10 Hz). Anal. ( C ~ ~ H ~ O F ~ ~ ~ O S Z ~ ~ ~ H Z ~ ) C, H, S.

Test Methods. Human Vasoconstrictor Activity. The method used was a modification of that described by McKenzie and Atkinson." Six male and six female subjects with six marked sites on the flexor surface of each forearm were treated with 12 test solutions, allocated by means of a 12 x 12 latin square. The test solutions (0.02 mL) were four graded doses of the standard fluocinolone acetonide (3,12.5,50,200 ng) and of two test steroids in ethanol. They were pipetted on to the marked circular areas (ca. 2.25 cm2) and spread as evenly as possible. The solvent was allowed to evaporate, and the forearms were enclosed in polythene tubing secured at each end by elastic surgical tape. The tubing was left in place overnight (ca. 16 h), and 1 h after its removal the arms were assessed for areas of vasoconstriction. The areas were scored as either positive or negative, and estimates of relative potency were obtained by analysis of the data using the method of Litchfield and Wilcoxon.28

Topical ihthflamma tory Activity in Rats. The method used was a modification of that described by Tonelli et al.lS and used 7 groups of 10 rats (54-70 g) in each experiment. Croton oil soluton was prepared by mixing croton oil (5 vol), EtOH (20 vol), and EbO (75 vol). Graded doses of the standard fluocinolone acetonide and test steroid were dissolved in the croton oil solution such that the doses (standard steroid 0.06, 0.25, and 1 pg) were contained in 0.04 mL. The doses were applied to the inner aspect of both ears, the control group receiving croton oil solution without medication. Six hours later, the rats were sacrificed by C02 inhalation and both ears were removed and weighed separately. The metameters used for the analyses of variance and calculation of relative potencies were the logarithm of the applied dose and ear weights.

Topical Antiinflammatory Activity in Mice. The methodology was similar to that for rats except that a mixture of croton oil (2 vol), EtOH (20 vol), and Et20 (78 vol) was used and the dose volume was 0.02 mL. Nine groups of 10 mice (22-28 g) were used, and doses of the standard fluocinolone acetonide were 0.015, 0.06, 0.25, and 1 pg.

Systemic Corticosteroid Activity after Topical Ap- plication to Rats. Seven groups of six rats (170-200 g at the beginning of the experiment) were used. The fur on the dorsal skin was removed by clipping, and a circular area (ca. 2 cm2) was marked using an indelible marker. Graded doses of the standard fluocinolone acetonide and test steroid were dissolved in acetone such that the doses (0.5, 2, 8 pg of standard) were contained in 0.02 mL. The rats were treated once daily for seven consecutive days; one group which received unmedicated vehicle acted as controls. On the day after the final dose, the rats were stressed by exposure to Et20 vapor for 1 min. Twenty minutes later, the rats were re-anaesthetised with intraperitoneal pentobarbitone, and blood was withdrawn by cardiac puncture into heparinized tubes. Plasma corticosterone levels were measured for each blood sample by a modification of the fluorimetric method of Zenker and Bern~tein. '~ Alternatively, adrenal glands were removed, blotted, and weighed, The metameters used for the analyses of variance and calculation of relative potencies were the

(0.068 g, 65.6%): 'H NMR (90 MHz) 0.99 (13-CH3, s), 1.06


(6) Bain, B. M.; May, P. J.; Phillipps, G. H.; Woollett, E. A. Anti- inflammatorv Esters of Steroidal Carboxvlic Acids. J . Steroid

logarithm of the daily dose and plasma corticosterone concentrations or adrenal weights.

Systemic Corticosteroid Activity after Topical Ap- plication to Mice. The methodology was similar to that for rats except that nine groups of 10 mice (25-30 g) were used and that, after stressing, the mice were re-anaesthetised with Et2O. Daily doses of the standard fluocinolone acetonide were 0.47, 1.9, 7.5, and 30 pug. The blood from two mice was pooled to obtain sufficient for the corticosterone estimation.

Systemic Corticosteroid Activity after Subcutaneous or Oral Administration to Rats and Mice. The methods involved seven daily doses of the standard betamethasone (2.5, 10, and 40 pg to rats and 1.56, 6.25, and 25 pg to mice) and the test steroid given subcutaneously as a suspension in normal saline containing 0.5% Tween 80 by volume (0.2 mL for rats and 0.1 mL for mice). Adrenal weights in rats and corticosterone levels in mice were measured as for topical application above. Oral activities were similarly assessed.

Acknowledgment. We thank Miss P, J. McDonough for the elemental analyses, Dr. R. A. Fletton and C. J. Seaman for the NMR and IR measurements, and Dr. S. J. Lane for the mass spectra. We are indebted to Dr. P. J. May for discussions and help in the preparation of the paper, and finally, but not least to Miss L. Micklewright for her patience and skill in preparing the manuscript.

Supplementary Material Available: 'H NhfR data (5 pages). Ordering information is given on any current mast- head page.

References (1) Phillipps, G. H. Locally Active Corticosteroids: Structure-

Activity Relationships. In Mechanisms of Topical Corticosteroid Activity; Wilson, L., Marks, R., Eds.; Churchill Livingstone: London, 1976; pp 1-18.

(2) Popper, T. L.; Watnick, A. S. Anti-inflammatory steroids. In Anti-inflammatorv Aeents. Chemistrv and Pharmacologv: Scher- rer, R. A., Whitehbu&, M.'W., Eds.; Academic Press: I&wYork, 1974; Vol. 1, pp 245-294.

(3) Elks, J.; Phillipps, G. H. Discovery of a Family of Potent Topical Anti-inflammatory Agents. In Medicinal Chemistry. The Role of Organic Chemistrv in Drug Research: Roberts. S. M., Price, B. J.,Eds.; Academic Press:-London, 1985; pp 167-188. We are grateful to a referee for the correct suggestion that the product formed from an androstane-178-carboxylic acid 3 by reaction with ClCSNMez is the (rearranged) mixed thioanhy- dride 4b, instead of the thione anhydride 4a postulated previously.

(4) Lee, H. J.; Taraporewala, I. B.; Heiman, A. S. Anti-inflammatory Steroids: Research Trends and New Compounds. Drugs Today 1989,25,577-588.

( 5 ) Part of this material was presented in preliminary form. Phillipps, G. H. Structure-Activity Relationships of Topically Active Steroids: the Selection of Fluticasone Propionate. Respir. Med. 1990, 84 (Supplement A), 19-23.

Biochem. l 9 i4 , 5, 299. (7) P h i l l i ~ ~ s . G. H.: Bain, B. M. US. Patent 4093721,1978; German

Pateit-2538595, 1974; Chem. Abstr. 1976,85, 63240d.

Abstr. 1975. 82. 31453~: British Patent 1384372, 1974. (8) Phillipps, G. H.; May, P. J. US. Patent 3828080, 1974; Chem.

Phillipps, G. H: Structure-Activity Relationships in Steroidal Anaesthetics, J . Steroid Biochem. 1975, 6, 607-613. Phillipps, G. H.; Marshall, D. R. US. Patent 3989686, 1980; Chem. Abstr. 1979, 86, 7300313. Kertesz, D. J.; Marx, M. Thiol Esters from Steroid 17/3-Carboxylic Acids: Carboxylate Activation and Internal Participation by 17a- Acylates. J. Org. Chem. 1986, 51, 2315-2328. Gais, H. J. Synthesis of Thiol and Selenol Esters from Carboxylic Acids and Thiols or Selenols, respectively. Angew. Chem., Znt. Ed. Engl. 1977,16, 244-246. Janssen, M. J. In Chemistry of Carboxylic Acids and Esters; Patai, S., Ed.; Interscience-Publishers, John Wiley & Sons Ltd: London, 1969; pp 705-764. Burton, D. J.; Greenlimb, P. E. Fluoro Olefins VII. Preparation of Terminal Vinyl Fluorides. J . Org. Chem. 1975, 40, 2796- 2810. Robinson, J. M. British Patent 2216122, 1989; German Patent Application P3906273.2,1989; Chem. Abstr. 1990,112,54961~. Klayman, D. L.; Griffin, T. S. Reaction of Selenium with Sodium Borohydride in Protic Solvents. A Facile Method for the Introduction of Selenium into Organic Molecules. J . Am. Chem.

L

SOC. 1973, 95, 197-199. McKenzie, A. W.; Atkinson, R. M. Topical Activities of Be- tamethasone Esters in Man. Arch. Dermatol. 1964, 89, 741- 746. Tonelli, G.; Thibault, L.; Ringler, I. A Bio-assay for the Con- comitant Assessment of the Antiphlogistic and Thymolytic Activities of Topically Applied Corticoids. Endocrinology 1965,

Zenker, N.; Bernstein, D. E. The Estimation of Small Amounts of Corticosterone in Rat Plasma. J . Biol. Chem. 1958,231,695- 701. Harding, S. M. The Human Pharmacology of Fluticasone Pro- pionate. Respir. Med. 1990, 84 (Supplement A), 25-29. Trademark of the Glaxo group of companies. Batres, E.; Bowers, A.; Djerassi, C.; Kinch, F. A.; Mancera, 0.; Ringold, H. J.; Rosenkranz, J.; Zaffaroni, A. German Patent 1079042,1960; Chem. Abstr. 1961,55, 15551e. Fried, J.; Florey, K.; Sabo, E. F.; Herz, J. E.; Restivo, A. R.; Borman, A.; Singer, F. M. Synthesis and Biological Activity of 1- and 6-Dehvdro-Sa-Halocorticoids. J . Am. Chem. SOC. 1955.

77,625-634.

77, 4181-41g2. Amello, E. J.; Laubach, G. D.; Moreland, W. T. US. Patent 3667197, 1962; Chem. Abstr. 1963,58, 9187b. Bain, B. M.; Woollett, E. A,; May, P. J.; Phillipps, G. H. British Patent 1438940, 1976; German Patent 2336633, 1974; Chem. Abstr. 1974, 80, 121208j. Clark, J . C.; Bain, B. M.; Phillipps, G. H. British Patent 1517278, 1974. Corey, E. J.; Schmidt, G. Useful Procedures for the Oxidation of Alcohols Involving Pyndinium Dichromate in Aprotic Media. Tetrahedron Lett. 1979, 399-402. Litchfield, J. T.; Wilcoxon, F. Simplified Method for Evaluating Dose-effect Experiments. J . Pharmacol. Exp. Ther. 1949, 96, 99-113.

V

Ha

b

c

a

ARA

KRIFRA

1

ginier[lt

mpsftti

smdf

1d

Spectrochimica Acta Part A 72 (2009) 244–247

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular andBiomolecular Spectroscopy

journa l homepage: www.e lsev ier .com/ locate /saa

ibrational spectroscopic study of fluticasone propionate

.R.H. Ali a,c, H.G.M. Edwardsa,∗, J. Kendrickb, I.J. Scowena

Chemical & Forensic Sciences, University Analytical Centre, School of Life Sciences, University of Bradford, Bradford, BD7 1DP, UKInstitute of Pharmaceutical Innovation, University of Bradford, Bradford, BD7 1DP, UKDepartment of Pharmaceutical Analytical Chemistry, Faculty of Pharmacy, University of Assiut, Assiut, Egypt

r t i c l e i n f o

rticle history:eceived 24 July 2008

a b s t r a c t

Fluticasone propionate is a synthetic glucocorticoid with potent anti-inflammatory activity that has beenused effectively in the treatment of chronic asthma. The present work reports a vibrational spectroscopic

ccepted 13 August 2008

eywords:aman spectroscopy

nfrared spectroscopyluticasone propionate

study of fluticasone propionate and gives proposed molecular assignments on the basis of ab initio calcu-lations using BLYP density functional theory with a 6-31G* basis set and vibrational frequencies predictedwithin the quasi-harmonic approximation. Several spectral features and band intensities are explained.This study generated a library of information that can be employed to aid the process monitoring offluticasone propionate.

aup

2

2

S

2

Bmic5L

espiratory pharmaceuticalsb initio structural calculations

. Introduction

Fluticasone propionate (FP) is a highly potent, triflourinatedlucocorticoid. Unlike most other corticosteroids, the structure oft is based on androstane, rather than a pregnane, corticosteroiducleus (Fig. 1). The molecule is designed to maximize topical anti-

nflammatory activity and minimise the unwanted systemic sideffects associated with other glucocorticoids; this is because FP isapidly metabolised to the inactive 17�-carboxylic acid metabolite1–3]. The combination of potent topical anti-inflammatory and aow systemic side effect profile makes FP an ideal compound toreat asthma, seasonal and perennial rhinitis [4,5].

Raman spectroscopy has proved to be a simple and reliableethod for the determination of the composition profile of solid

harmaceutical samples [6–10]. Due to its non-invasiveness, highensitivity and good reproducibility, in addition to the requirementor little or no sample preparation, this technique is an importantool for the screening of drugs, once the unique fingerprint spec-rum that is specific for each pharmaceutical compound has beendentified.

In this paper, we present the first vibrational spectroscopic
tudy of FP and propose a vibrational assignment of the experi-ental spectral features aided by ab initio calculations using BLYP
ensity functional theory with a 6-31G* basis set and vibrationalrequencies and intensities predicted within the quasi-harmonic

∗ Corresponding author.E-mail address: [email protected] (H.G.M. Edwards).

2

Sts

c

386-1425/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.saa.2008.08.004

© 2008 Elsevier B.V. All rights reserved.

pproximation. The results thus obtained permit a rapid andnequivocal spectroscopic identification of FP in pharmaceuticalreparation and processes.

. Experimental

.1. Materials

A specimen of fluticasone propionate was obtained from theigma–Aldrich Chemical Co. and used without further purification.

.2. Raman spectroscopy

Fourier-transform Raman spectroscopy was carried out using aruker IFS 66 instrument with an FRA 106 Raman module attach-ent and a Nd3+/YAG laser operating at 1064 nm in the near

nfrared. The powdered specimens were examined in aluminiumups. The spectra were recorded at 4 cm−1 spectral resolution and00 spectral scans accumulated to improve signal-to-noise ratios.aser powers were maintained at 100 mW at the sample.

.3. Infrared spectroscopy

The IR spectra were recorded as KBr discs (1:200) using a Digilabcimitar 2000 Series spectrometer. The spectra were recorded overhe range of 650–4000 cm−1 at 4 cm−1 spectral resolution and 512pectral scan accumulations.

Raman and infrared spectra were analysed using a GRAMS7urve resolution package.

http://www.sciencedirect.com/science/journal/13861425

http://www.elsevier.com/locate/saa

mailto:[email protected]

dx.doi.org/10.1016/j.saa.2008.08.004

H.R.H. Ali et al. / Spectrochimica Acta Part A 72 (2009) 244–247 245

F1S

2

ststttadn

3

w4u1am(plbdi

F

F

RFsRcitncc

sosftRaRmcm2

ig. 1. The chemical structure of fluticasone propionate (androsta-1,4-diene-7-carbothioic acid, 6,9-difluoro-11-hydroxy-16-methyl-3-oxo-17-(1-oxopropoxy)-(fluoromethyl) ester (6�,11�,16�,17�).

.4. Calculation details

Calculations were performed using GAMESS-UK [11] and ORCAoftware packages [12]. An initial structure (code DAXYUX) forhe geometry optimisation was taken from the Cambridge Crystaltructure Database (CSD) optimised using BLYP density functionalheory [13,14]. Infrared and Raman spectra were calculated usinghe quasi-harmonic approximation, through diagonalisation ofhe mass-weighted Hessian matrix. Intensities for the infrarednd Raman transitions were calculated using the dipole-momenterivatives and the polarisability derivatives respectively of theormal modes.

. Results and discussion

The infrared and Raman spectra of fluticasone propionateere recorded over the wavenumber range 4000–600 cm−1 and000–100 cm−1, respectively, using both transmission and atten-ated total reflectance modes of operation in the former and a064 nm laser wavelength of excitation in the latter. The infrarednd FT-Raman spectra are shown in Figs. 2–4. A complete assign-ent of the experimental vibrational features was carried out

Table 1) in the light of both the theoretical results presently
erformed and the vibrational spectroscopic data. Several calcu-
ated modes have been matched with the experimentally observedands in the IR and Raman spectra, and these modes have beenescribed. The vibrational spectra illustrate clearly the complex-

ty of structural information that is provided from the infrared and

ig. 2. FT-Raman and IR spectral stackplot of FP in the 2600–3700 cm−1 region.

3

iEtb1maTtatatsai1p1


aman analysis; the OH stretching modes observed near 3500 cm−1

ig. 2 in the infrared spectrum are not observed in the Ramanpectrum, whereas bands in the low wavenumber region of theaman spectrum that are normally characteristic of polymorphicrystalline pharmaceutical forms are not recorded in the lowernfrared region Fig. 4. Although the infrared and Raman spec-ra have been reported previously [15,16], these reports provideeither spectral band assignments nor vibrational spectroscopicharacterisation in conjunction with the molecular ab initio cal-ulations.

The profile of the �(OH) stretching band (Fig. 2) in the infraredpectrum at 3336 cm−1 suggests that external hydrogen bondingccurs in the crystalline FP. The CH stretching region compriseseveral features in the wavenumber range 3100–2800 cm−1 (Fig. 2or the infared and Raman spectra); the CH stretching bands ofhe unsaturated, component of FP are shown more clearly in theaman spectrum rather than in the infrared spectrum (Table 1)nd can be assigned to the bands at 3075 and 3058 cm−1 in theaman spectrum in addition to, the CH stretching band of the five-embered ring at 3023 cm−1. The CH aliphatic stretching bands

omprise both CH3 and CH2 symmetric and asymmetric stretchingodes which have been assigned to the medium features between

878 and 2965 cm−1.

.1. 1800–1550 cm−1 region

This is a very important spectroscopic region for structural stud-es, particularly for the C O and C C groups in the molecule.xpanded wavenumber regions of the infrared and Raman spec-ra are shown in Fig. 3. The infrared spectrum has one very strongand at 1661 cm−1 and four strong bands at 1744, 1699,1616 and611 cm−1 compared to one very strong band at 1663 cm−1, oneedium band at 1606 cm−1, medium shoulder band at 1611 cm−1

nd two weak bands at 1742 and 1698 cm−1 in the Raman spectrum.he strong IR bands at 1744 and 1699 cm−1 in the IR spectrum andhe weak Raman spectral feature at 1742 and 1698 cm−1 can bettributed to the carbonyl attached to the aliphatic ring (cf. litera-ure values for a five-membered aliphatic ring at 1750–1700 cm−1,nd dimethyl ketone at 1715 cm−1) [17] and carbonyl attachedo sulphur, respectively, this shift occurs because of force con-
tant changes in the carbonyl bond [18], whereas that at 1661nd 1616 cm−1 can be attributed to the C O and C C stretch-ng vibration in the quinonoid aromatic ring (cf. benzophenone at660 cm−1), which couples together in the same moiety. This cou-ling explains the observed strength in both Raman and infrared at663 and 1661 cm−1, respectively.

246 H.R.H. Ali et al. / Spectrochimica Acta Part A 72 (2009) 244–247

Table 1The observed and calculated vibrational wavenumbers/cm−1 of fluticasone propionate. The calculated Raman and infrared intensities (IRaman and IIR) are relative to that of�(C O) at 1667 cm−1, set arbitrarily equal to 1.

Observed Calculated Proposed assignment

Raman (cm−1) IR (cm−1) � IRaman IIR

3336 m 3567 0.513 0.030 �(OH)3075 w 3069 vw 3123 0.186 0.035 �(CH)3058 mw 3051 vw 3082 0.465 0.069 �(CH)3023 mw 3024 vw 3027 0.229 0.120 �(CH) of the five-membered ring3005 mw 2975 m 2999 0.291 0.119 �(C CH)2965 mw 2964 m 2968 0.138 0.080 �as (CH3), �as (CH2)2935 m 2942 m 2944 0.308 0.155 �as (CH2)2878 mw 2881 m 2933 0.356 0.102 �s (CH2)1742 w 1744 s 1737 0.684 0.047 �(C O)1698 w 1699 s 1695 0.118 0.525 �(S C O)1663 vs 1661 vs 1667 1.000 1.000 �(C O)

1616 s 1639 0.036 0.155 �(C C)1611 sh,m 1611 s1606 m 1607 0.054 0.039 ı(CH3)1492 mw 1530 vw 1493 0.071 0.019 ı(CH3)1453 mw 1455 m 1445 0.027 0.058 ı(CH3)1422 w 1421 mw 1425 0.004 0.024 Ring stretching1387 w 1395 mw 1399 0.041 0.017 ı(CH2)1369 w 1376 mw 1372 0.028 0.0451334 w 1365 mw 1357 0.019 0.002 ı(CH)1321 m 1319 m 1323 0.074 0.0341281 w 1304 m 1298 0.081 0.059 �(C C)1233 mw 1271 m 1267 0.035 0.074 �(C F)

1248 m 1244 0.018 0.0551194 mw 1226 m 1223 0.097 0.052 In-plane ı(CH)

1214 m 1217 0.015 0.0161168 mw 1188 m 1177 0.017 0.0221146 mw 1147 w 1147 0.004 0.3971126 w 1124 m 1124 0.167 0.0461106 w 1109 0.021 0.0161067 w 1064 s 1070 0.039 0.0901058 w 1058 0.018 0.0551023 w 1027 m 1020 0.023 0.480 �(F C S)992 w 993 s 993 0.097 0.196 r(CH3)970 w 972 0.004 0.053 r(CH3)953 w 944 m 953 0.047 0.257 ı(C C C)931 w 933 m 936 0.015 0.124887 w 895 ms 904 0.028 0.135 OOC/CCH aromatic deformation

884 ms 877 0.011 0.094 OOC/CCH aromatic deformation869 w 873 m 863 0.016 0.110 OOC/CCH aromatic deformation843 w 835 0.033 0.028 �(O C O C)

816 w 816 0.022 0.046799 vw 797 w 802 0.002 0.014773 vw 782 w 776 0.021 0.032 r(CH2)733 m 734 mw 751 0.008 0.009698 vw 707 w 706 0.003 0.013 CH wagging650 vw 655 0.005 0.037 �(C S)616 vw 626 0.015 0.012 �(C S)584 w 581 0.005 0.016562 w 558 0.034 0.015549 w 541 0.004 0.084532 w 524 0.022 0.009519 w 512 0.019 0.012 Out-of-plane ı(CC)505 w 508 0.002 0.014470 vw 484 0.019 0.016448 vw 448 0.008 0.004435 vw 425 0.009 0.007409 w 408 0.013 0.033387 vw 387 0.014 0.018370 w 372 0.015 0.025348 vw 348 0.013 0.014330 vw 324 0.007 0.203308 w 311 0.019 0.008286 vw 282 0.011 0.08264 w 264 9.6e-4 0.018 CH3�234 w 237 0.013 0.026 CH3�159 w 156 0.010 0.011

H.R.H. Ali et al. / Spectrochimica Act

F

3

stiso1dt1svsito

9ooa

toeIsn

bomi

4

oioic

A

t

R

[

[

[

[[[

[


.2. 1800–100 cm−1 region

Infrared and Raman spectra over this wavenumber range arehown in Fig. 4; the infrared region comprises a rich spectrum fromhe C O and C C functionalities, which has been already discussedn some detail above, along with several bands of medium andtrong intensities, whereas the Raman spectrum consists mainlyf the strong C C features and other weaker features. In the range530–1330 cm−1, where we expect CH3, CH2 and CH aliphaticeformation vibration to occur. Lower in wavenumber, we expecto find the aliphatic C F stretching vibrations at 1271 cm−1 and233 cm−1, of medium intensity in the infrared and of weakertrength in the Raman spectrum, respectively. The in-plane ı(CH)ibration occurs at 1226 cm −1 in the IR spectrum of medium inten-ity and at 1194 cm−1 in the Raman spectrum of medium weakntensity and the �(F C S) band occurs at 1027 cm−1 in the IR spec-rum of medium intensity and at 1023 cm−1in the Raman spectrumf weak intensity.

Methyl rocking and CH2 rocking modes extend over the range70–995 cm−1, the OOC/CCH aromatic deformation modes extendver the range 870–900 cm−1, the (O C O C) stretching modeccurs at 843 cm−1, the CH wagging mode occurs near 700 cm−1

nd the C S stretching mode occurs in the range of 616–650 cm−1.Finally, the complex ring modes of CCO and COCOC vibra-

ions and deformations in the range between 600 and 500 cm−1

ccur in the low wavenumber region; in this region we shouldxpect the aromatic quinonoid ring deformations near 600 cm−1.n the Raman spectrum below 400 cm−1, the methyl tor-ional mode at 234 and 264 cm−1 of weak intensity can beoted.

[

[

a Part A 72 (2009) 244–247 247

Even with the support of ab initio calculations, it is difficult toe certain about some molecular vibrational assignments becausef the complexities of mode mixing; however, it is clear that theajor spectroscopic features of FP have been reasonably assigned

n this study.

. Conclusions

A complete vibrational spectroscopic analysis has been carriedut for fluticasone propionate. Raman, infrared and molecular ab

nitio calculations, have been used to explain the presence of thebserved spectroscopic complexity in the C C and C O stretch-ng region which will provide useful structural information whenonsidering the complexation of fluticasone propionate.

cknowledgement

Hassan R.H. Ali is grateful for the support of the Government ofhe Arab Republic of Egypt during which this work was carried out.

eferences

[1] S.C. Sweetman (Ed.), Martindale the Extra Pharmacoepia, 34th ed., The Phar-maceutical Press, London, 2005.

[2] G.H. Phillipps, E.J. Bailey, B.M. Bain, R.A. Borella, J.B. Buckton, J.C. Clark, A.E.Doherty, A.F. English, H. Fazarkerley, S.B. Laing, E. Lane-Allman, J.D. Robinson,P.E. Sandford, P.J. Sharratt, I.P. Steeples, R.D. Stonehouse, C. Williamson, J. Med.Chem. 37 (1994) 3717.

[3] B.M. Bain, G. Harrison, K.D. Jenkins, A.J. Pateman, E.V.B. Shenoy, J. Pharm.Biomed. Anal. 11 (1993) 557.

[4] R. Fuller, M. Johnson, A. Bye, Respir. Med. A89 (1995) 3.[5] G.K. Scadding, V.J. Lund, L.A. Jacques, D.H. Richards, Clin. Exp. Allergy 25 (1995)

737.[6] S.E.J. Bell, D.T. Burns, A.C. Dennis, J.S. Speers, Analyst 126 (2001) 541.[7] S.E.J. Bell, D.T. Burns, A.C. Dennis, L.J. Matchett, J.S. Speers, Analyst 125 (2000)

1811.[8] B. Sagmuller, B. Schwarze, G. Brehm, S. Schneider, Analyst 126 (2001) 2066.[9] K. Faulds, W.E. Smith, D. Graham, R.J. Lacey, Analyst 127 (2002) 282.10] T. Vankeirsbilck, A. Vercauteren, W. Baeyens, G. Van der Weken, F. Verpoort, G.

Vergote, J.P. Remon, Trends Anal. Chem. 21 (2002) 869.11] M.F. Guest, I.J. Bush, H.J.J. Van Dam, P. Sherwood, J.M.H. Thomas, J.H. Van Lenthe,

R.W.A. Havenith, J. Kendrick, Mol. Phys. 103 (2005) 719.12] F. Neese, ORCA—an ab initio, density functional and semiempirical program

package, version 2.4, revision 45, Max Planck Institut für BioanorganischeChemie, Mülheim and der Ruhr (2005).

13] A.D. Becke, Phys. Rev. A 38 (1988) 3098.14] C.T. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.15] Y. Guo-feng, G. Xiao-chong, G. Rui-chang, Y. Zhao-xun, A. Zhi-qiang, Hecheng

Huaxue 15 (2007) 510.16] A. Theophilus, A. Moore, D. Prime, S. Rossomanno, B. Whitcher, H. Chrystn, Int.

J. Pharm. 313 (2006) 14.17] N.B. Colthup, L.H. Daly, S.E. Wiberley, Introduction to Infrared and Raman Spec-

troscopy, Academic Press, London, 1975.18] D. Lin-Vien, N.B. Colthup, W.G. Fateley, J.G. Grasselli, The Handbook of Infrared

and Raman Characteristic Frequencies of Organic Molecules, Academic Press,London, 1991.

Improved Synthesis of Fluticasone PropionateJiadi Zhou, Can Jin, and Weike Su*

Key Laboratory for Green Pharmaceutical Technologies and Related Equipment of Ministry of Education, Collaborative InnovationCenter of Yangtze River Delta Region Green Pharmaceuticals, College of Pharmaceutical Sciences, Zhejiang University ofTechnology, Hangzhou 310014, P. R. China

*S Supporting Information

ABSTRACT: A novel process for the preparation of fluticasone propionate (1), a corticosteroid, is reported. In this paper,compound 2 was used as starting material to prepare 6 by using NaClO or NaBrO which was much cheaper than H5IO6 as anoxidizing agent. Furthermore, toxic, expensive, and pollutive BrCH2F was replaced by AgNO3 and Selectfluor in decarboxylativefluorination.

■ INTRODUCTION

Steroidal glucocorticoid agonists such as fluticasone propionate(1) are anti-inflammatory agents used widely against a broadspectrum of inflammatory diseases. Fluticasone propionate (1),synthesized by Glaxo Wellcome and launched in 1993, is atrifluorinated glucocorticosteroid. It shows good topical anti-inflammatory activity and is commonly used as a safe andeffective inhaled treatment for asthma and allergic rhinitis.1 Theprevious route for the synthesis of 6 from 2 was in thecommercial scale. Compound 6 was synthesized fromcommercial grade flumethasone (5) by H5IO6 oxidation witha yield of 95.0%.2a Flumethasone (5) can be prepared from 2 inthree steps with a unclear yield (Scheme 1).3 According to theliterature,2a,3 we obtained 6 from 2 in a total yield of 45%.The compound 7 was synthesized from 6 by propionyl

chloride or propionic anhydride acylation. Compound 7reacted with N, N-dimethylthiocarbamoyl chloride4 in thepresence of an iodide catalyst and Et3N to produce 8, followedby hydrolyzation with K2CO3,

2a Et2NH,2b or NaSH4a to obtain

9. Alternatively, the combination of 1,1′-carbonyldiimidazole(CDI) with NaSH2b can also be used to prepare 9 from 7.Fluticasone propionate (1) can be synthesized from 9 by usingBrCH2F,

2a ClCH2F,5a or S-(monofluoromethyl) diarylsulfo-

nium tetrafluoroborate5f,g directly. Using BrCH2F can get anideal yield; however, BrCH2F is costly and will destroy to theozone layer. In addition, 9 reacted with BrCH2Cl or Br2CH2and then by an anion exchange with AgF,2c,5e KF, ortetrabutylammonium fluoride5b to afford 1 in a low yield.Fluticasone propionate (1) could be obtained from 10, in thepresence of fluorodecarboxylating reagents such as XeF2 andBrF3.

5d Unfortunately, XeF2 is extremely expensive, and BrF3which should be stored in Teflon containers is a strongcorrosive toxic liquid, which tends to react very exothermicallywith water and release poisonous vapours. Furthermore, thehigh toxicities and instabilities of XeF2 and BrF3 preventedpractical applications of this method. According to theliterature,5c Deoxo-Fluor or DAST can also be used asmonofluoromethylation reagent to acquire 1 from 11 at −60°C (Scheme 2). Considering the different literature sources, thehighest overall yield for the previous synthesis of fluticasonepropionate (1) from 2 was close to 30%.

The disadvantages of the above processes include safetyissues, high expenses, and environmental problems, such as theuse of costly H5IO6 as an oxidant and BrCH2F, XeF2, BrF3, orDeoxo-Fluor as monofluoromethylation reagents. Consideringthese drawbacks, we have subjected this synthetic route tofurther researches and intended to develop an efficient, eco-friendly, and commercially feasible process for fluticasonepropionate (1). In this article, we describe an improved processwith an overall yield of 42.3% in method A and 54.5% inmethod B (Scheme 3).

■ RESULTS AND DISCUSSIONPreparation of 6α,9α-Difluoro-11β,17α-dihydroxy-

16α-methyl-3-oxoandrosta-1,4-diene-17β-carboxylicAcid (6). As one of the oldest known organic reactions, thehaloform reaction can be used to convert a terminal methylketone into appropriate carboxylic acid.6 Applying this reactionto the synthesis of compound 6 from 2 could shorten reactionroutes,7 reduce the cost,8 and improve the total yield comparedwith the traditional method in four steps. Luckily, we foundthat using NaClO in the presence of NaOH could successfullyobtain compound 6 in room temperature with a yield of63.7%,9 and 11β,17α-dihydroxy and 1,4-diene were tolerated. Aseries of solvents such as dioxane/H2O, THF/H2O, dimethoxy-ethane/H2O, and EtOH/H2O, were screened in order to findan optimal solvent. The results showed that all of those mixedsolvents did not work well to afford the desired product 6except THF/H2O/EtOH, which provided a homogeneousreaction system.By contrast, the usage of NaBrO which has a stronger activity

than NaClO could get a higher yield of 84.5% in a lowertemperature (Scheme 3). To this reaction, dioxane/H2O wasproved to be the best solvent.10 Other solvents such as THF/H2O, dimethoxyethane/H2O, EtOH/H2O, and THF/H2O/EtOH did not work well, and the reaction did not proceed inbiphasic systems. When the reaction was finished, theremaining oxidizing agent (NaClO or NaBrO) was destroyedby the addition of excess sodium sulfite solution, and the

Received: April 14, 2014Published: July 24, 2014

Article

pubs.acs.org/OPRD

© 2014 American Chemical Society 928 dx.doi.org/10.1021/op5001226 | Org. Process Res. Dev. 2014, 18, 928−933

pubs.acs.org/OPRD

solvent THF/EtOH or dioxane was removed under reducedpressure. After extraction with ethyl acetate, the aqueous phasewas acidified with hydrochloric acid to furnish a whiteprecipitate of 6, which was collected by filtration, washedwith water, and dried.Preparation of 6α,9α-Difluoro-11β-hydroxy-16α-

methyl-17α-propionyloxy-3-oxoandrosta-1,4-diene-17β-carbothioate (10). According to the literature,2a,5d thecompound 10 can be synthesized from 6 in four steps includingesterification, acylation, alcoholysis, and alkylation (Scheme 3).In the original methods,5d the product 10 was acquired from 9by BrCH2COOH alkylation under the condition of using DCMas the solvent. In our improved process, lower toxic solventacetone was used to replace DCM. When the reaction wasfinished, the compound 10 could obtained by filtrationconveniently after acidification with 1 mol/L HCl.Preparation of S-Fluoromethyl-6α,9α-difluoro-11β-

hydroxy-16α-methyl-17α-propionyloxy-3-oxoandrosta-1,4-diene-17β-carbothioate (1). Recently reported N−F

reagents, such as Selectfluor and NFSI, are commerciallyavailable, easy to use, and stable electrophilic fluorinatingreagents that can be used to conduct decarboxylativefluorination.11 In our improved process, fluticasone propionate(1) was prepared from 10 with AgNO3/Selectfluor (Scheme4).12 According to the literature,11a the combination of AgNO3

(20 mol %) with Selectfluor (2.5 equiv) shows a considerabledecarboxylative fluorination ability (Table 1, entry 1). OtherAg(I) salts, such as AgOAc and AgOTf, exhibited a weakercatalytic activity (Table 1, entries 2 and 3), while no reactionoccurred without the presence of a Ag(I) salt (Table 1, entry4). Switching the electrophilic fluorinating reagent fromSelectfluor to NFSI caused no reaction (Table 1, entry 5). Inaddition to Ag(I) ions, water also was turned out to be essential(Table 1, entry 6). Much of the experimental results was similarto the optimization done by the Li group.11a

The Ag(II)- or Ag(III)-mediated decarboxylation ofcarboxylic acids is well-documented.13 According to theliterature,11a a tentative mechanism of the decarboxylative

Scheme 1. Previous route for the synthesis of 6

Scheme 2. Previous method for the synthesis of fluticasone propionate (1)a

aReactions and conditions: (a) (i) propionic anhydride or propionyl fluoride/Et3N; (ii) Et2NH; (b) N,N-dimethylthiocarbamoyl chloride/Et3N/NaIor tetrabutylammonium iodide; (c) K2CO3 or Et2NH or NaSH; (d) CDI/NaSH; (e) BrCH2COOH/Et3N; (f) formaldehyde; (g) XeF2/BrF3; (h)BrCH2F or ClCH2F or S-(monofluoromethyl) diarylsulfonium tetrafluoroborate or BrCH2Cl/AgF or BrCH2Cl/KI/KF; (i) Deoxo-Fluor or DAST.

Organic Process Research & Development Article

dx.doi.org/10.1021/op5001226 | Org. Process Res. Dev. 2014, 18, 928−933929

fluorination with AgNO3/Selectfluor was proposed. Theoxidation of Ag(I) by Selectfluor generates an Ag(III)−Fintermediate, which initiated the decarboxylative fluorination ofcarboxylic acids (Figure 1).Further optimization of the decarboxylative fluorination was

carried out by screening of a range of temperatures, and ofreagent stoichiometries, using AgNO3 and Selectfluor (Table2). As shown in (Table 2, entry 1), 10 was treated with AgNO3(20 mol %) and Selectfluor (2.5 equiv) at 30 °C for 8 h under anitrogen atmosphere giving 1 in 80.1% yield, when raising thetemperature to 45 °C improved the yield of 1 to 92.7% (Table2, entry 2). However, more impurities were generated when the

reaction was performed at 55 °C (Table 2, entry 3).Unfortunately, using AgNO3 (10 mol %)/Selectfluor (2.5equiv) or AgNO3 (20 mol %)/Selectfluor (2.0 equiv) asfluorodecarboxylating reagents provided 1 in only 75.2% and70.3% yields, respectively (Table 2, entries 4 and 5).The activity of decarboxylative fluorination was proved to be

solvent-dependent, the mixed solvent of acetone/H2O (2:1,v:v) exhibited the best activity (Table 2, entry 2). By contrast, ahigher reaction temperature was needed in CH3CN/H2O (2:1,v:v) solution that led to more impurity 12 (Table 2, entry 6).No fluorodecarboxylation occurred in THF/H2O (2:1, v:v)solution (Table 2, entry 7) or other biphasic systems.The optimized reaction conditions (Table 2, entry 2), 10 was

treated with AgNO3 (20 mol %) and Selectfluor (2.5 equiv) inacetone/H2O (2:1, v:v) solution at 45 °C for 3 h under anitrogen atmosphere giving the expected product fluticasonepropionate 1. Dilution with water upon completion of thereaction and collection of the product by filtration followed bywashing with water gave 1 in 92.7% isolated yield.The major impurity 12 was obtained by preparative HPLC,

the process of impurity formation was studied (Scheme 5). Wesuspected that fluticasone propionate (1) could be hydrolyzed

Scheme 3. Improved process for the synthesis of fluticasone propionate (1)a

aReactions and conditions: (a) NaClO/NaOH, THF/EtOH/H2O, 25 °C, 63.7%; (b) NaBrO/NaOH, dioxane/H2O, 0−5 °C, 84.5%; (c) (i)propionic anhydride/Et3N, acetone, 15−25 °C; (ii) Et2NH, 15−25 °C, 97.3%; (d) N,N-dimethylthiocarbamoyl chloride/Et3N/NaI, acetone/H2O,30 °C, 96.0%; (e) K2CO3, CH3OH, 25 °C, 95.0%; (f) BrCH2COOH/Et3N, acetone, 15−25 °C, 97.1%; (g) AgNO3/Selectfluor, acetone/H2O, 45°C, 92.7%.

Scheme 4. Decarboxylative fluorination for the synthesis of 1

Table 1. Screening of Ag(I) catalyst and N−F reagent ofdecarboxylative fluorination of 10a

entrycatalyst(equiv)

N−Freagent(equiv) solvent

time(h)

yieldb

(%)purityc

(%)

1 AgNO3 Selectfluor acetone/H2O 3 92.7 92.62 AgOAc Selectfluor acetone/H2O 8 64.2 74.23 AgOTf Selectfluor acetone/H2O 8 48.8 79.54 Selectfluor acetone/H2O 8 05 AgNO3 NFSI acetone/H2O 8 06 AgNO3 Selectfluor acetone 8 trace

aReagents and conditions: 1.0 equiv 10, 0.2 equiv catalyst, 2.5 equivN−F reagent, 45 °C, under nitrogen. bThe isolated yield wascalculated with 10. cThe purity was monitored by HPLC.

Figure 1. Proposed mechanism of silver-catalyzed decarboxylativefluorination.



to 7 which might generate 12 through decarboxylativefluorination rapidly. The C-17 stereochemistry of compound12 was established by X-ray crystallography. In order to confirmthe speculation, we successfully using 7 as material to obtain 12with AgNO3/Selectfluor in acetone/H2O solution at 45 °C.

■ CONCLUSIONIn conclusion, an efficient, eco-friendly, and commerciallyviable process for the synthesis of fluticasone propionate (1)has been developed. In this method, compound 2 was used asthe starting material, which was transformed to 1 in six stepsincluding oxidation, esterification, acylation, alcoholysis,alkylation, and fluorodecarboxylation. Especially, compared totraditional flumethasone oxidation with H5IO6, application ofhaloform reaction to the synthesis of compound 6 for the firsttime could shorten reaction routes, reduce the cost, andimprove the total yield dramatically. Furthermore, the usage oftoxic, extremely costly, and pollutive BrCH2F was replaced byan efficient method with AgNO3 and Selectfluor in a goodyield.

■ EXPERIMENTAL SECTIONCompound 2 was provided by Zhejiang Xianju JunyePharmaceutical Co., Ltd., and all other chemicals werepurchased from commercial sources and were used withoutfurther purification. HPLC analysis for fluticasone propionate(1) was carried out on an Agilent HPLC system (series 1200,Agilent Technologies, Germany) equipped with AgilentZORBAX SB-C18 reversed-phase column (250 mm × 4.6mm, 5 μm). A mobile phase of methanol, acetonitrile, andbuffer with 1.2 g/L of monobasic ammonium phosphate, a pHof 3.5 adjusted with phosphoric acid, (50:15:35) was used at aflow rate of 1.5 mL/min and a column temperature of 40 °C.

The UV detector was set at 239 nm to analyze the columneffluent. 1H (400 MHz) NMR, 13C (101 MHz) NMR, and 19F(376 MHz) NMR spectra were recorded on a Varianspectrometer in CDCl3 or DMSO-d6 using tetramethylsilane(TMS) as internal standards.

6α,9α-Difluoro-11β,17α-dihydroxy-16α-methyl-3-ox-oandrosta-1,4-diene-17β-carboxylic Acid (6). Method A.NaOH (25.0 g, 0.625 mol) was dissolved in H2O (0.040 L), themixture was diluted with 0.500 L of THF and 0.500L of EtOH.Into the solution, 2 (50.0 g, 0.126 mol) was added, then 10%NaClO (0.500 L) was added gradually at 25 °C. The reactionwas kept at 25−30 °C for 6 h. When the reaction was finished,the remaining oxidizing agent (NaClO) was destroyed by theaddition of excess 10% Na2SO3 solution. The solvent THF/EtOH was removed under reduced pressure. Ethyl acetate(0.500 L) was added to the solution; the layers were separated,and then acidification of the aqueous phase to pH = 1.0−2.0 by3 mol/L HCl furnished a white precipitate of 6 (32.0 g), whichwas collected by filtration, washed with water, and dried. Yield:63.7%; HPLC purity 96.0%.

Method B. Br2 (140.0 g, 0.875 mol) was added slowly to avigorously stirred solution of 130.0 g of NaOH in 1.170 L ofH2O while cooling in an ice-salt-bath. When all of the Br2 haddissolved, the mixture was diluted with 0.600 L of cold dioxane,and the ice-cold NaBrO was added slowly to a stirred solutionof 100.0 g of 2 in 1.400 L of dioxane which was maintained at atemperature below 8 °C throughout the oxidation. After 5 h,the remaining oxidizing agent (NaBrO) was destroyed by theaddition of excess 10% Na2SO3 solution. The solvent dioxanewas removed under reduced pressure. Ethyl acetate (1.000 L)was added to the solution; the layers were separated, thenacidification of the aqueous phase to pH = 1.0−2.0 by 3 mol/LHCl furnished a white precipitate of 6 (85.0 g), which wascollected by filtration, washed with water, and dried. Yield:84.5%; HPLC purity 95.6%; 1H NMR (400 MHz, DMSO-d6) δ12.45 (s, 1H), 7.24 (d, J = 10.3 Hz, 1H), 6.27 (dd, J1 = 10.2 Hz,J2 = 1.6 Hz, 1H), 6.08 (s, 1H), 5.70−5.54 (2m, 1H), 5.32 (s,1H), 4.69 (s, 1H), 4.12 (d, J = 11.7 Hz, 1H), 2.88−2.80 (m,1H), 2.50−2.35 (m, 2H), 2.25−1.96 (m, 3H), 1.70−1.50 (m,2H), 1.49 (s, 3H), 1.12−1.05 (m, 1H), 0.99 (s, 3H), 0.86 (d, J= 7.0 Hz, 3H); 13C NMR (101 MHz, DMSO-d6) δ 184.14 (s),174.35 (s), 162.86 (d, J = 13.7 Hz), 151.86 (s), 128.85 (s),119.23 (d, J = 12.9 Hz), 100.13 (d, J = 176.0 Hz), 86.83 (d, J =178.0 Hz), 85.34 (s), 70.61 (d, J = 36.0 Hz), 48.14 (d, J = 19.5Hz), 47.31 (s), 42.23 (s), 35.42 (s), 35.19 (s), 33.91 (d, J = 18.8Hz), 32.30 (m), 31.99 (s), 22.86 (s), 16.93 (s), 15.45 (s);MS(ESI-) m/z 395.2 [M − H]+.

6α,9α-Difluoro-11β-hydroxy-16α-methyl-17α-propio-nyloxy-3-oxoandrosta-1,4-diene-17β-carboxylic Acid(7). To a suspension of 6 (85.0 g, 0.215 mol) in acetone(0.425 L) at 10−15 °C was added sequentially Et3N (65.1 g,

Table 2. Optimisation of decarboxylative fluorination of 10

entry AgNO3 (equiv) Selectfluor (equiv) solvent temp (°C) time (h) yielda (%) purityb (%)

1 0.2 2.5 acetone/H2O 30 8 80.1 85.02 0.2 2.5 acetone/H2O 45 3 92.7 92.63 0.2 2.5 acetone/H2O 55 3 92.0 79.74 0.1 2.5 acetone/H2O 45 8 75.2 80.25 0.2 2.0 acetone/H2O 45 8 70.3 83.16 0.2 2.5 MeCN/H2O 55 8 80.5 67.07 0.2 2.5 THF/H2O 55 8 0

aThe isolated yield was calculated with 10. bThe purity was monitored by HPLC.

Scheme 5. Process of impurity 12 generated



0.645 mol) and propionic anhydride (83.8 g, 0.645 mol). Afterstirring for 4 h at 25 °C, Et2NH (31.4 g, 0.430 mol) was addeddropwise at 10−15 °C and then stirred at 25 °C for 1 h.Thereafter, the reaction mixture was acidified to pH 1.0−1.5with 1 mol/L HCl at 0 °C. The precipitated product 7 (93.3 g)was obtained by filtered, washed with water, and dried. Yield96.0%; HPLC purity 97.0%; 1H NMR (400 MHz, DMSO-d6) δ7.24 (d, J = 10.3 Hz, 1H), 6.27 (dd, J = 10.1, 1.9 Hz, 1H), 6.09(s, 1H), 5.70−5.54 (2m, 1H), 5.46 (s, 1H), 4.16 (m, 1H), 3.14(m, 1H), 2.50 (m, 1H), 2.30 (q, J = 7.2 Hz, 2H), 2.23 (m, 1H),2.05 (m, 2H), 1.82−1.68 (m, 2H), 1.49 (m, 4H), 1.18 (m, 1H),1.01 (t, J = 7.2 Hz, 3H), 1.00 (s, 3H), 0.84 (d, J = 7.2 Hz, 3H);13C NMR (101 MHz, DMSO-d6) δ 184.03 (s), 171.96 (s),169.79 (s), 162.59 (d, J = 13.5 Hz), 151.64 (s), 128.88 (s),119.24 (d, J = 12.1 Hz), 99.96 (d, J = 175.9 Hz), 91.13 (s),86.69 (d, J = 180.8 Hz), 70.23 (d, J = 36.1 Hz), 47.96 (d, J =22.2 Hz), 47.64 (s), 42.63 (s), 35.41 (s), 35.30 (s), 33.86 (d, J =18.9 Hz), 33.08 (s), 32.19 (m), 27.02 (s), 22.77 (s), 16.44 (s),16.43 (s), 9.26 (s); MS(ESI+) m/z 475.4 [M + Na]+.

6α,9α-Difluoro-11β-hydroxy-16α-methyl-17α-propio-nyloxy-3-oxoandrosta-1,4-diene-17β-carbothioic Acid(9). A solution of 7 (93.3 g, 0.206 mol) and N,N-dimethylthiocarbamoyl chloride (50.8 g, 0.449 mol) in acetone(1.866 L) at room temperature was cooled to 10−15 °C. It wassequentially treated with Et3N (41.3 g, 0.413 mol), NaI (15.0 g,0.080 mol), and water (9.330 mL, 10% w/w with 7) at 10−15°C. The solution was stirred for 6 h at 30 °C, then added DMF(0.466 L) and water (3.000 L). The resultant was cooled to 0°C and stirred for 1 h. The precipitated product 8 (106.6 g)was obtained by filtration, washed with water, and dried. Yield96.0%; HPLC purity 96.5%.A suspension of 8 (106.6 g, 0.196 mol) and K2CO3 (54.1 g,

0.392 mol) in methanol (0.530 L) was stirred at 25 °C for 5 hunder a blanket of nitrogen. Thereafter, water (0.530 L) wasadded to the reaction mixture, and the resultant clear solutionwas washed twice with toluene (0.212 L). The aqueous layerwas acidified with 1 mol/L HCl until pH is 1.5 to 2.0. Theprecipitated product was filtered, washed with water, and driedto obtain 9 (87.1 g). Yield 95.0%; HPLC purity 96.0%; 1HNMR (400 MHz, DMSO-d6) δ 7.25 (d, J = 11.2 Hz, 1H), 6.30(dd, J = 10.1, 1.8 Hz, 1H), 6.11 (s, 1H), 5.80 (d, J = 5.0 Hz,1H), 5.72−5.55 (2m, 1H), 4.27 (m, 1H), 3.30 (m, 1H), 2.64−2.54 (m, 1H), 2.40 (q, J = 7.5 Hz, 2H), 2.30−2.09 (m, 4H),1.92−1.88 (m, 1H), 1.51 (m, 4H), 1.26 (m, 1H), 1.13 (s, 3H),1.03 (t, J = 7.5 Hz, 3H), 0.87 (d, J = 6.9 Hz, 3H); 13C NMR(101 MHz, DMSO-d6) δ 189.76 (s), 184.00 (s), 172.36 (s),162.45 (d, J = 13.5 Hz), 151.38 (d, J = 10.9 Hz), 128.92 (s),119.27 (d, J = 12.5 Hz), 99.73 (d, J = 176.4 Hz), 96.90 (s),86.63 (d, J = 174.0 Hz), 69.85 (d, J = 36.1 Hz), 48.34 (s), 47.87(d, J = 24.4 Hz), 42.90 (s), 36.63 (s), 35.44 (s), 33.76 (d, J =18.8 Hz), 33.51 (s), 32.01 (m), 27.07 (s), 22.84 (s), 17.08 (s),15.51 (s), 9.05 (s); MS(ESI−) m/z 467.1 [M − H]+.6α,9α-Difluoro-11β-hydroxy-16α-methyl-17α-propio-

nyloxy-3-oxoandrosta-1,4-diene-17β-carbothioate (10).A solution of 9 (87.1 g, 0.186 mol), Et3N (28.2 g, 0.279 mol),and BrCH2COOH (28.2 g, 0.205 mol) in acetone (0.871 L)was stirred at 25 °C for 5 h. Thereafter, water (0.871 L) wasadded, and the reaction mixture was acidified to pH 1.0−1.5with 1 mol/L HCl at 0 °C. The precipitated product 10 (95.0g) was obtained by filtered, washed with water, and dried. Yield97.1%; HPLC purity 97.0%; 1H NMR (400 MHz, DMSO-d6) δ7.24 (d, J = 10.6 Hz, 1H), 6.28 (dd, J = 10.0, 1.6 Hz, 1H), 6.10(s, 1H), 5.60 (d, J = 4.4 Hz, 1H), 5.70−5.53 (2m, 1H), 4.19

(m, 1H), 3.67 (s, 2H), 3.25 (m, 1H), 2.56−2.45 (m, 1H), 2.33(q, J = 7.4 Hz, 2H), 2.24 (m, 1H), 2.08 (m, 2H), 1.91−1.86 (m,2H), 1.48 (m, 4H), 1.24 (m, 1H), 1.02 (s, 3H), 1.00 (t, J = 7.6Hz, 3H), 0.89 (d, J = 6.9 Hz, 3H); 13C NMR (101 MHz,DMSO-d6) δ 194.93 (s), 184.01 (s), 171.81 (s), 169.19 (s),162.49 (d, J = 13.4 Hz), 151.46 (d, J = 8.1 Hz), 128.90 (s),119.25 (d, J = 12.8 Hz), 99.84 (d, J = 176.4 Hz), 95.63 (s),86.68 (d, J = 180.7 Hz), 70.00 (d, J = 35.0 Hz), 48.65 (s), 47.90(d, J = 19.9 Hz), 42.71 (s), 35.75 (s), 35.07 (s), 33.78 (d, J =19.3 Hz), 33.40 (s), 31.98 (m), 31.47 (s), 27.09 (s), 22.77 (s),17.03 (s), 15.83 (s), 9.10 (s); MS(ESI−) m/z 525.0 [M − H]+.

S-Fluoromethyl-6α,9α-difluoro-11β-hydroxy-16α-methyl-17α-propionyloxy-3-oxoandrosta-1,4-diene-17β-carbothioate (1). A solution of 10 (95.0 g, 0.180 mol),Selectfluor (159.3 g, 0.450 mol), and AgNO3 (6.1 g, 0.036 mol)in acetone (1.900 L) and water (0.950 L) was stirred at 45 °Cfor 3 h under a blanket of nitrogen. Then water (1.900 L) wasadded to the solution; the resultant was cooled to 0 °C andstirred for 1 h. The precipitated product 1 (83.5 g) wascollected by filtered, washed with water, and dried. Yield 92.7%;HPLC purity 92.6%.

Purification. The crude product 1 (83.5 g) was dissolved inethyl acetate (0.835 L) and ethanol (3.340 L). The suspensionwas refluxed for 30 min, gradually cooled to 0 °C, and stirredfor 1 h, and the soild was collected by filtration and dried at 40°C under vacuum to provide 69.5 g (83%) of product 1. HPLCpurity 99.2%; residual silver content 0.04633 (μg/g).14

1H NMR (400 MHz, DMSO-d6) δ 7.23 (d, J = 10.0 Hz, 1H),6.28 (dd, J = 10.1, 1.6 Hz, 1H), 6.10 (s, 1H), 5.92 (J = 50.0 Hz,2H), 5.70−5.54 (2m, 1H), 5.58 (d, J = 3.2 Hz, 1H), 4.20 (m,1H), 3.28 (m, 1H), 2.36 (q, J = 7.2 Hz, 2H), 2.23 (m, 1H), 2.09(m, 2H), 1.86 (m, 2H), 1.53 (m, 1H), 1.48 (s, 3H), 1.26 (m,1H), 1.05 (m, 1H), 1.01 (t, J = 7.2 Hz, 3H), 0.99 (s, 3H), 0.89(d, J = 7.1 Hz, 3H); 13C NMR (101 MHz, DMSO-d6) δ 192.87(s), 183.98 (s), 172.07 (s), 162.43 (d, J = 13.5 Hz), 151.54 (s),128.91 (s), 119.25 (d, J = 12.1 Hz), 99.72 (d, J = 176.3 Hz),95.94 (s), 86.62 (d, J = 178.0 Hz), 80.92 (d, J = 211.8 Hz),69.97 (d, J = 37.2 Hz), 48.40 (s), 47.84 (d, J = 22.4 Hz), 42.84(s), 35.74 (s), 35.10 (s), 33.73 (d, J = 19.4 Hz), 33.37 (s), 31.93(m), 26.94 (s), 22.73 (s), 16.95 (s), 16.08 (s), 9.05 (s); 19FNMR (376 MHz, CDCl3) δ −165.35 (dd, J = 27.5, 8.5 Hz),−187.00 (dd, J = 48.3, 13.8 Hz), −191.35 (t, J = 49.6 Hz);MS(ESI+) m/z 501.0 [M + H]+.

6α,9α,17α-Trifluoro-11β-hydroxy-16α-methyl-17β-propionyloxy-3-oxoandrosta-1,4-diene (12). 1H NMR(400 MHz, CDCl3) δ 7.11 (d, J = 10.1 Hz, 1H), 6.41 (s,1H), 6.35 (dd, J = 10.1, 1.2 Hz, 1H), 5.45−5.29 (2m, 1H), 4.35(m, 1H), 2.58−2.46 (m, 1H), 2.35 (q, J = 7.6 Hz, 2H), 2.28−2.23 (m, 1H), 2.05−1.68 (m, 6H), 1.52 (s, 3H), 1.35−1.28 (m,1H), 1.18−1.07 (m, 9H); 13C NMR (101 MHz, CDCl3) δ185.20 (s), 171.28 (s), 160.94 (d, J = 13.8 Hz), 150.23 (s),130.10 (s), 121.93 (d, J = 250 Hz), 121.08 (d, J = 9.7 Hz),98.53 (d, J = 177.4 Hz), 86.47 (d, J = 181.1 Hz), 71.76 (d, J =35.1 Hz), 48.05 (d, J = 22.6 Hz), 47.08 (d, J = 20.6 Hz), 42.09(s), 39.58 (s), 39.36 (s), 37.79 (s), 33.21 (m), 32.08 (s), 28.13(s), 23.25 (s), 16.44 (s), 15.32 (d, J = 13.1 Hz), 9.04 (s); 19FNMR (376 MHz, CDCl3) δ −129.60 (d, J = 17.3 Hz), −165.39(dd, J = 27.5, 8.3 Hz), −187.05 (dd, J = 48.3, 13.7 Hz);HRMS(ESI+): C23H30F3O4 [M + H]+; calculated: 427.2091,found: 427.2089.



christinehurd

Highlight

■ ASSOCIATED CONTENT

*S Supporting Information

Safety assessment for the NaClO oxidation and fluorodecarbox-ylation, 1H and 13C NMR copies for compounds 6, 7, 9, 10, 1,and 12, 19F NMR copies for compounds 1 and 12, and X-raycrystallographic data for compound 12. This material isavailable free of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATION

Corresponding Author*Telephone/fax: (+86)57188320899. E-mail: [email protected].

Notes

The authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

We gratefully acknowledge financial support from NationalNatural Science Foundation of China (No. 21176222) andCollaborative Innovation Center of Yangtze River Delta RegionGreen Pharmaceuticals.

■ REFERENCES(1) (a) Barnes, P. J. Pharmaceuticals 2010, 3, 514−540. (b) Rhen, T.;Cidlowski, J. A. New Engl. J. Med. 2005, 353, 1711−1723.(2) (a) Jadav, K. J.; Kambhampati, S.; Chitturi, T. R.; Thennati, R.World Patent WO 2004/001369, 2004. (b) Phillipps, G. H.; Bailey, E.J.; Borella, R. A.; Buckton, J. B.; Clark, J. C.; Doherty, A. E.; English, A.F.; Fazakerley, H.; Laing, S. B.; Lane-Allman, E.; Robinson, J. D.;Sandford, P. E.; Sharratt, P. J.; Steeples, I. P.; Stonehouse, R. D.;Williamson, C. J. Med. Chem. 1994, 37, 3717−3729. (c) Myhren, F.;Sandvold, M. L.; Eriksen, O. H.; Hagen, S. World Patent WO 2008/115069 A2, 2008.(3) Wu, M. Z. China Patent CN 101838301 A, 2010.(4) (a) Barkalow, J.; Chamberlin, S. A.; Cooper, A. J. World PatentWO 01/62722 A2, 2001. (b) Qin, G. R. China Patent CN100497367C, 2009.(5) (a) Partridge, J. J.; Walker, D. S. World Patent WO 03/013427A2, 2003. (b) Shen, Y. L.; Liu, X. L.; Xie, L. B.; He, H. X. China PatentCN 100549022C, 2009. (c) Cainelli, G.; Umani, R. A.; Sandri, S.;Contento, M.; Fortunato, G. World Patent WO 2004/052912 A1,2004. (d) Leitao, E. P. T.; Heggie, W. World Patent WO 2011/151625A1, 2011. (e) Zhu, D. J.; Zhang, D. F. China Patent CN 100560598C2009. (f) Surya Prakash, G. K.; Ledneczki, I.; Chacko, S.; Olah, G. A.Org. Lett. 2008, 10, 557−560. (g) Leitao, E. P. T.; Turner, C. R. WorldPatent WO 2011/151624 A1, 2011.(6) Djerassi, C.; Staunton, J. J. Am. Chem. Soc. 1961, 83, 736−743.(7) Compound 13, acquired by chemical degradation of Saponin, is abasic raw material for the synthesis of steroidal glucocorticoid drugs.Compound 14 (17-keto), acquired by biological selective degradationphytosterol side chain, is a basic raw material for the synthesis of sexhormones. It is difficult to synthesize 21-desoxy from 17-keto.Compound 2 could be prepared from 13 in several steps, andflumethasone (5) was prepared from 2. In our improved process,applying haloform reaction to the synthesis of compound 6 from 2could shorten reaction routes, reduce the cost, and improve the totalyield (see scheme below).

(8) We have checked the prices for NaClO, Br2, and H5IO6 on atonne scale. 10% NaClO (96 $ per ton), Br2 (288 $ per ton), andNaBrO could be prepared by Br2 on the spot. H5IO6 needs 136085 $per ton, and it was not friendly to the environment.(9) The yield of NaClO oxidation (63.7%) was higher than previoussynthesis in four steps (45%), even if it was lower than NaBrOoxidation (84.5%). Furthermore, the NaClO oxidation system wasmore environmentally friendly and suitable for industrialization.(10) Dioxane could be recycled under reduced pressure, and thewater in the recycled dioxane has no significant influence for the nexthalogen reaction.(11) (a) Yin, F.; Wang, Z. T.; Li, Z. D.; Li, C. Z. J. Am. Chem. Soc.2012, 134, 10401−10404. (b) Ruedabecerril, M.; Sazepin, C. C.;Leung, J. C. T.; Okbinoglu, T.; Kennepohl, P.; Paquin, J. F.; Sammis,G. M. J. Am. Chem. Soc. 2012, 134, 4026−4029. (c) Leung, J. C.;Chatalova-Sazepin, C.; West, J. G.; Rueda-Becerril, M.; Paquin, J. F.;Sammis, G. M. Angew. Chem., Int. Ed. 2012, 51, 10804−10807.(d) Mizuta, S.; Stenhagen, I. S. R.; O’Duill, M.; Wolstenhulme, J.;Kirjavainen, A. K.; Forsback, S. J.; Tredwell, M.; Sandford, G.; Moore,P. R.; Huiban, M.; Luthra, S. K.; Passchier, J.; Solin, O.; Gouverneur,V. Org. Lett. 2013, 15, 2648−2651. (e) Rueda-Becerril, M.; Mahe, O.;Drouin, M.; Majewski, M. B.; West, J. G.; Wolf, M. O.; Sammis, G. M.;Paquin, J. F. J. Am. Chem. Soc. 2014, 136, 2637−2642.(12) In the previous routes, BrCH2F is used as the majorelectrophilic monofluoromethylation reagent. However, BrCH2F iscostly (2500 $ per kg) and toxic and will destroy the ozone layer. Thecost of Selectfluor (170 $ per kg)/AgNO3 (500 $ per kg) is lower thanBrCH2F. Furthermore, Selectfluor/AgNO3 is more environmentallyfriendly than BrCH2F.(13) (a) Anderson, J. M.; Kochi, J. K. J. Am. Chem. Soc. 1970, 92,1651−1659. (b) Anderson, J. M.; Kochi, J. K. J. Org. Chem. 1970, 35,986−989. (c) Kouadio, I.; Kirschenbaum, L. J.; Mehrotra, R. N.; Sun,Y. J. Chem. Soc., Perkin Trans. 2 1990, 12, 2123−2127.(14) According to the ICH Q3D, we calculated the USP InhalationLimit of Silver was close to 0.69 (μg/g). The residual silver content ofcompound 1 was 0.04633 (μg/g) by ICP-MS detection. In addition,the silver salts in the aqueous waste stream could be recycled. Silversalts reacted with HCl or NaCl could generate the AgCl precipitationwhich could be collected by filtration.



http://pubs.acs.org



J. Org. Chem. 1986,51,2315-2328 2315

Thiol Esters from Steroid 17P-Carboxylic Acids: Carboxylate Activation and Internal Participation by 17a-Acylates'

Denis J. Kertesz* and Michael MBrx

Syntex Research, Palo Alto, California 94304

Received September 24, 1985

The chemistry of the steroid 17@-carboxylic acids derived from 16,17a-disubstituted corticosteroids was investigated with respect to thiol ester formation. Major quantities of 17-spiro byproducts were observed in the reactions of 16-methyl-17a-acyloxy acids, and the degree of 17-ester participation leading to these structures was dependent on the carboxylate activating group used and stereochemistry at C-16. Diethyl phosphate mixed anhydrides of these acids reacted with mercaptide salts to give mixtures of thiol esters with 17-spiro acylthio ortho esters, which predominated and were particularly stable in the case of l6@-methyl substrates; in addition, considerable reversion of l6a-methyl phosphate intermediates to starting acid wm experienced. The use of diphenyl chlorophosphate as the activating agent greatly improved yields of thiol esters. Methanolysis of the phosphate adducts derived from l7a-acyloxy acids gave 17-spiro acyl ortho esters as the exclusive products. The reactions of 17a-acetoxy acids with 2-fluoro-N-methylpyridinium tosylate (FMPT) gave the novel 17-spiro acyl fluoro ketals 32-35, whereas similar treatment of 17-hydroxy acids led to products of dehydration or of 18-methyl migration, including the novel 13,17-P-lactones 39 and 41. Activation with carbonyldiimidazole followed by addition of mercaptans allowed the preparation of thiol ester products from 17-hydroxy acids, but the method was restricted to use with these substrates. Neighboring-group participation was not possible for the 16,17-acetonide acid 10, and activation with either chlorophosphate diesters or FMPT followed by reaction with methanethiolate gave high yields of methylthio ester 17.

Introduction Dermatologically useful antiinflammatory activity in

corticosteroids is compatible with a variety of modifications to the normal 20-keto-21-hydroxy side chain, as exemplified by the 21-deoxy, 21-deoxy-21-chloro, and steroid l7P-carboxylate ester classes of topical agents.2 In the course of our investigations of structure-activity relationships in this area, we required the synthesis of thiol esters of steroid 17P-carboxylic (etianic) acids bearing either the 16a,l7m-acetonide functionality or a 16-methyl in combination with a 17a-acyloxy or -hydroxy group.3 Although the thiol esters of 16-unsubstituted-17-desox- yetianic acids were k n o ~ n , ~ , ~ there was no precedent for synthesis of the analogous 16,17a-disubstituted derivatives. The system of interest is particularly hindered, the 170- carboxyl being flanked by a quaternary carbon (C-13), a tertiary carbon (C-16), and a l7a-oxy substituent. These features had a profound influence on the chemistry of activation and further reactivity of the 17@-carboxylate, particularly through participation by the neighboring 17a-acyloxy and -hydroxy groups, and led to the isolation of a novel series of 17-spiro dioxolanones and 18-methyl migration products. The elaboration of these findings is reported here.

Carboxylic acid starting materials for the desired thiol esters were prepared from fluocinolone acetonide 1, flumethasone 2, betamethasone 3, the A9J1-16P-methyl corticoid 4, and the Sa,llp-dichloride 5 by oxidative cleavage of the corticoid side chain, usually followed by esterification of free 17a-alcohol groups (Scheme I). Thus, oxidation of the 17a-hydroxy substrates 2-5 with periodic acid6 in aqueous methanol gave high yields of etianic acids

(1) Contribution No. 704 from the Institute of Organic Chemistry. (2) (a) Wolff, M. E. In Burger's Medicinal Chemistry, 4th ed., Part

III; Wolff, M. E., Ed.; Wiley-Interscience: New York, 1981; pp. 1310-1311. (b) Phillips, G. H. In Mechanisms of Topical Corticosteroid Activity; Wilson, C.; Marks, R., Ed.; Churchill Livingstone: New York, 1976; pp 1-18.

(3) Edwards, J. A. US. Patent 4 188 835. 1980, (4) Jeger, 0.; Norymberski, J.; Szpilfogel, S.; Prelog, V. Helu. Chim.

(5) Phillips, G. H.; Marshall, D. R. US. Patent 3989686, 1980. (6) Reichstein, T.; Meystre, Ch.; von Euw, J. Helu. Chim. Acta 1939,

Acta 1946, 29, 684.

22. 1107.

0022-3263/86/1951-2315$01.50/0

6-9. The 16,17-acetonide 1 (and corticoid 17-esters) did not react under these conditions, showing an unexpected limitation to the use of periodic acid. A potassium carbonate catalyzed air oxidation was therefore developed based on reports by Velluz' and Herzig,8 permitting cleavage of 1 to the 16,17-acetonide acid 10 in 75% yield. The l7a-hydroxyetianic acids 6,7, and 9 were transformed into the correspondkg pcetates and propionates 11-15 in a one-pot procedure similar to that of Phillips et LLI . ,~JO consisting of reaction with excess acetyl or propionyl chloride and triethylamine followed by the addition of diisopropylamine to cleave the resulting 17-acyloxy carboxylic acid mixed phydrides. The mixed anhydrides of 16P-methyl acids were particularly stable, as demonstrated by the isolation of the dipropionyloxy compound 169 in 72% yield from 7 when the use of diisopropylamine was eliminated. The ease with which 17-esters could be prepared from etianic acid substrates suggests that the free acid group was involved in the process,*l probably through the initial formation of a mixed anhydride and transfer of the acyl group from the 17P-carboxylate to the 17a- alcohol.

Activation with Chlorophosphate Reagents. Thiol esters of 16,17-acetonide acid 10 were successfidly prepared by forming diethyl phosphate intermediated2 and treatment in situ with sodium mercaptide salts. Thus, reaction of 10 with diethyl chlorophosphate and triethylamine in THF followed by filtration of the amine hydrochloride salt

(7) Velluz L.; Petit, A.; Peaez, M.; Berret, R. Bull. SOC. Chim. Fr. 1947,

(8) Herzig, P. Th.;$hrenstein, M. J. Org. Chem. 1951, 16, 1050. (9) Phillips, G. H.; BGn, B. M. US. Patent 4093721, 1978. (10) Phillips, G. H.; May, J. M. US. Patent 3 828 080, 1974; Chem.

Abstr. 1975, 82, 31453~. (11) By contrast, the esterification of free 17-alcohols in etianic acid

thiol esters required catalysis with DMAP a t elevated temperatures, as in the preparations of 48 and 49 from 47 (vide infra). If an lip-alcohol was present, it was also esterified under these conditions. Selective base hydrolysis of the 11-esters of the resulting 11,l'l-diesters was readily achieved with I6a-methyl substrates, but no distinction was possible in the presence of a I6Smethyl group: Alvarez, F.; Kertesz, D., unpublished results.

(12) (a) Masamune, S.; Kampta, S.; DiaKur, J.; Sugihara, Y.; Bates, G. S. Can. J. Chem. 1975,53,3693. (b) For a review of thiol ester forming methods, see: Hadam, E. Tetrahedron 1980,36, 2409.

123; Chem. Abstr. 19 7, 41, 5537.

0 1986 American Chemical Society

2316 J. Org. Chem., Vol. 51, No. 12, 1986

Scheme I. Preparation of Etianic Acids

C H,OH

CH,OH

F

CH,OH

F

COOH

6: 6 , g - F ~ ; 11-OH; 16a-CH2

7: 9-F; ll-OH;lGB-CH3 8: 6-F;A9’11;16B-CH3 9: 6-F;9,11-C12;16B-CH3

- - - -

and addition of excess sodium methylthiolate in DMF gave 72% of methylthio ester 17 and 18% of unreacted 10, whereas the use of sodium ethanethiolate gave 65% of 18 (Scheme 11). However, application of this method to 16a-methyl-17a-acyloxy substrates gave uniformly low yields of the desired thiol esters accompanied by large quantities of recovered start ing acids. For example, acetate 11 gave methylthio ester 19 (13%) and ethylthio ester 20 (26%) accompanied in each case by 46-48% of starting acid, while 17-propionate 12 afforded 33% of methylthio ester 21 and 25% of returned 12.13

The reactions of etianic acids with chlorophosphate reagents produced single intermediates which could be observed by thin-layer chromatography (TLC), and when

(13) These results were representative of a series of reactions employing mercaptide8 as large as n-hexylthio in combination with 17- acylates as large as caproate.

Kertesz and Marx

COOH

YOOH

FOOH

FOOH

F

“70 ! C,H,

the intermediate formed from 11 (assumed h be phosphate 22) was titrated to its disappearance with NaSCH,, 12% of a new less polar material was formed in conjunction with 9% of thiol ester 19 and 34% of returne&acid 11. The new product showed a single absorption at 1790 cm-’ in the infrared in place of the 17-acylate and thiol ester peaks, and its identification as the novel 17-spiro acylthio orthoacetate 23 was supported by further spectr$l and analytical data. Although 23 was readily isomerFed to 19 with excess thiolate and could be hydrolyzed to 11 with either acid Gr base, the isolated material was stable under neutral conditions. The formation of 23 by preferential attack of thiolate at the 17-acyloxy carbonyl has analogy in the chemistry of a-acetoxyisobutyryl chloride studied by Mattocks14 and others,15 and this form of neighboring-

(14) Mattocks, A. R. J. Chem. Sot. 1964, 1918.

Thiol Esters from Steroid 17P-Carboxylic Acids J. Org. Chem., Vol. 51, No. 12, 1986 2317

Scheme 11. Thiol Esters via Phosphoric Mixed Anhydrides

COOH

+

19 - + 11 -

group participation was noted with increasing frequency in further work.

When we applied the diethyl phosphate method of thiol ester formation to 17-acyloxy substrates in the leb-methyl series, we found 17-spiro thio ortho esters to be the major products; the desired thiol esters were produced in low yields and amounts of recovered starting acids were in- significant. Moreover, chromatographic separation of the normal thiol ester and 17-spiro products derived from

lower alkyl thiolates was difficult or impossible. Thus, treatment of the phosphate intermediate derived from 17-acetate 13 with NaSC6H13 gave 43% of spiro hexylthio orthoacetate 28 and 13% of the readily separable thiol ester 25, whereas the reaction of similarly activated 17- propionate 14 with a stoichiometric amount of NaSCH, yielded 64% of an unresolvable mixture judged by IFt and NMR16 to be 3:l of orthopropionate 29 with ester 26. The

(15) (a) Greenberg, S.; Moffatt, J. G. J. Am. Chem. SOC. 1973,95,4016. (b) Alkoxydioxolanones (cyclic acyl ortho esters) are formed in the reactions of a-acetoxyisobutyryl chloride with alcohols. Treatment of this reagent with ethanethiol gave a mixture absorbing at 1800 cm-' in the infrared, implying the formation of an alkylthiodioxolanone similar to 23, but the product could not be isolated RQchardt, Ch.; Brinkmann, H. Chem. Ber. 1975,108, 3224.

(16) In the absence of adequate guidance from TLC, comparison of the intensity of the 1790 cm-' peak with that of the 17-ester carbonyl at 1730 cm-' was useful for judging the composition of ortho ester-thiol ester product mixtures. A large downfield shift of the 18-methyl resonance (0.30.35 ppm from that of the thiol eater) was the most dramatic feature of cyclic ortho ester NMR spectra, and other differences were used to calculate proportions in isomeric mixtures: see the Experimental Section.

2318 J. Org. Chem., Vol. 51, No. 12, 1986 Kertesz and Marx

2 2

Figure 1.

2 2 - I

19 -

+

Lt - conversion of ortho ester 29 to thiol ester 26 was completed by warming with an excess of NaSCH,, as followed by disappearance of the 1790 cm-' infrared absorption. Re- action of the diethyl phosphate adduct of acetate 13 with one equivalent of NaSCH, at 0 "C gave an estimated 9:l mixture of cyclic orthoacetate 27 with thiol ester 24, and a sample of pure 27 was obtained by painstaking chromatographic separation. The tendency of the 16/3-methyl group to enhance the formation of 17-spiro thio ortho esters and increase their stability toward excess thiolate was a major feature of the chemistry of this series."

Substitution of diphenyl chlorophosphate for the diethyl reagent markedly altered the course of subsequent reactions with thiolates; notably, reversion of intermediates derived from 16a-methyl acids to starting material did not occur and ortho ester formation was greatly reduced for 16Smethyl substrates, leading to increased yields of thiol esters in all cases. Thus, treatment of 16a-methyl propionate 12 with diphenyl chlorophosphate and TEA in dry THF for 1 h at 55 OC gave a single TLC visible intermediate, and subsequent addition of NaSCH, afforded an 89% isolated yield of 21. Similarly, the reaction of the diphenyl phosphate adduct of 160-methyl- 17-propionate 14 with an equivalent of NaSCH, gave a mixture of more than 1O: l of thiol ester 26 with ortho ester 29, and was readily driven on with excess thiolate and heat to give a 64% yield of pure 26. Acetate 13 likewise gave 74% of 24. The use of diphenyl chlorophosphate for activation of 10 gave a 75% of acetonide thiol ester 17.

The thiol ester forming reactions of lea-methyl acids activated with diethyl chlorophosphate were noteworthy for the considerable reversion to starting material experienced, and an attempt was made to relate this behavior to the nature of the actual intermediate. The similar quantities of thiol ester and returned acid obtained from reactions of thiolate with these diethyl phosphates could be rationalized in several ways (Figure 1). One possibility is that the true intermediates are synimetric anhydrides, resulting from a reported tendency of diethyl phosphate mixed anhydrides to "disproportioliate"l2* (path 1). Re- version of the mixed phosphate anhydride to carboxylic acid might also result from a partitioning of the intermediate by attack of thiolate on phosphorusla or the phos-

(17) Facilitated access of thiolate to the l7a-acyloxy carbonyl and shielding of the resulting dioxolanone product carbonyl by the 16@,18- dimethyl system probably both contribute to this finding.

Jz

phate ethoxy a-carbons21 (path 2) in preference to the 170-carboxylate and 17a-acyloxy carbonyl carbons (path 3). Either the absence of alkoxy groups or a lower susceptibility of phosphorus to thiolate attack could then explain the superiority of diphenyl phosphate for thiol ester synthesis in this series. It was also considered possible that displacement of phosphate by chloride might occur during the activation procedure, with the resulting acid chlorides being the true intermediates in some of the phosphate-mediated chemistry. The possibility of acid chloride intermediates was discounted by investigation of the salt precipitates filtered from the intermediate-forming reactions, since formation of the presumed mixed anhydrides from 11 and 13 using either diethyl or diphenyl chlorophosphate was accompanied in each case by the virtual quantitative recovery (relative to carboxylic acid) of chloride ion as pure triethylammonium chloride. We found that acid chlorides could actually be formed and used without i~olation,2~ and the reactions of 11 and 14 with thionyl chloride and TEA for 10 min at 0 OC followed by addition of NaSCH, thereby afforded 32 and 34% yields of respective methylthio esters 19 and 26. Only small amounts of starting acids were recovered, and there were no detectable 17-spiro thio ortho ester products from these reactions.

An attempt was made to isolate and characterize a representative set of phosphate intermediates. Thus, reaction mixtures containing the 16a-methyl-17-acetoxy acid 11 and the 16B-methyl analogue 13 in combination with both diethyl and diphenyl chlorophosphate reagents were prepared as usual and subjected to aqueous workup pro-

(18) Nucleophilic attack on phosphorus was suggested by Liu et al. to explain low yields of eaters from the use of chlorophosphate diesters as activating agents (see ref 19). Low yields were also reported for thiol ester preparations using these reagents (see ref 20), but no difference between diethyl and diphenyl phosphate cases was noted in the cited work.

(19) Liu, H. 3.; Chan, W. H.; Lee, S. P. Tetrahedron Lett. 1978,4461. (20) Liu, H. J.; Sabesan, S. I. Can. J . Chem. 1980, 58, 2645. (21) Monodealkylation of the diethyl phosphate intermediate by so-

dium mercaptide (see ref 22) would give an anhydride salt, which could lead to free carboxylic acid by loss of metaphosphate or hydrolysis in the workup (see ref 23).

(22) Savignac, P.; Lavielle, G. Bull. Chim. SOC. Fr. 1974, 1506. (23) Moon, M. W.; Khorana, H. G. J. Am. Chem. SOC. 1966,88,1798. (24) The reactions of 11 and 14 with thionyl chloride gave TLC visible

intermediates with very similar R j s to those noted in reactions with diphenyl chlorophosphate, which at first suggested that acid chlorides might play a role in this chemistry.

Thiol Esters from Steroid 17@-Carboxylic Acids J. Org. Chem., Vol. 51, No. 12, 1986 2319

Scheme 111. Methanolysis of Phosphate-Activated 17a-Acetoxy Acids

n Q

111,

""C H3

cedures. Only the 16a-methyl diethyl chlorophosphate product 22 proved stable enough to be isolated in this

&22 '", CH,

2 2 - /

F

manner, with starting acid alone being isolated from the other reaction mixtures. Intermediate 22 was isolated in 84% yield; it survived purification by chromatography, and its identity was the desired diethyl phosphate mixed anhydride was conf i ied by spectral and analytical data. To our surprise in a series of TLC experiments with mercaptide salts, the isolated phosphate 22 gave erratic yields of less than 10% of substitution products with NaSCH,, these being roughly equal amounts of thiol ester 19 and spiro ortho ester 23, and virtually no such products were obtained with NaSC2H,. In both cases the free acid 11 was the single major product. The inconsistent behavior of 22 toward thiolate before and after isolation is puzzling, but it is clear that reversion to free acid is a characteristic of the diethyl phosphate species itself, and is not a result of "disproportionation".

Although phosphate-activated etianic acids were found to be unreactive toward free mercaptans,PS the solvolysis of these species with methanol proceeded with ease to give high yields of 17-spiro products. The products were those of 17-ester participation exclusively; no normal methyl ester products were detected. Thus, the 17-spiro acyl orthoacetate 30 was isolated in 71% yield from a solution of isolated phosphate 22 in methanol after heating at reflux for 1 h (Scheme 111). Isolation of phosphate intermediates was not necessary to produce methoxydioxolanones, and the exchange of methanol for the THF in freshly filtered reaction mixtures containing either diethyl or diphenyl phosphate anhydrides and warming gave excellent yields of 17-spiro products in all cases. The 16a- and 16-methyl orthoacetates 30 and 31 were prepared in this manner from 11 and 13 via the diethyl phosphates in 85% and 76%

(25) Carboxylate intermediates formed with phenyl dichlorophosphate are reported to be especially reactive toward free thiols (see ref 201, but the use of this reagent offered no advantage in the present work.

yields, respectively. Diethyl phosphates gave somewhat better yields than the diphenyl intermediates, and acid chloride preparations gave the same methanolysis products, but in poorer yields than phosphates.

The thiol esters of base-sensitive substrates and of free 17-hydroxy acids were inaccessable from phosphate intermediates. Thus, an attempted preparation of thiol ester 49 from 15 failed due to sensitivity of the 9,ll-dichloride system to the required thiolate, and reactions of 17- hydroxy acids 6 and 7 with chlorophosphate reagents gave only polar materials presumed to be 17-phosphate esters.

Reactions with 2-Fluoro-N-methylpyridinium To- sylate. In view of the limitations of chlorophosphate reagents, the use of 2-fluoro-N-methylpyridinium tosylateB (FMPT) for etianic acid activation was investigated. However, FMPT-derived intermediates did not react with free mercaptans, and neighboring group participation again dominated the chemistry. Thus, although the reaction of acetonide acid 10 with freshly prepared FMPT and TEA at -15 OC followed by addition of NaSCH, afforded a 92% yield of thiol ester 17, the chemistry of substrates bearing l7a-acyloxy or -hydroxy groups was complex. Treatment of the 16a-methyl-17-acetoxyetianic acid 11 with FMPT and TEA gave complete conversion to a mixture of two compounds of similar Rf that were totally unreactive toward a large excess of NaSCH,, even after 16 h at 110 OC. The products, isolated in a total yield of 64% by preparative TLC, were identified as the separate epimers 32 (34%) and 33 (30%) of a 17-spiro acylfluoro ketaI (Scheme IV). Both cyclic ketals absorbed at 1805 cm-l in the infrared and are related in structure and presumably in mechanism of formation to the previously noted acyl- orthoesters, but the complete incorporation of fluoride ion in this manner2' and the stability of the resulting fluoroketal structures were quite unexpected. Proton NMR spectroscopy allowed the assignment of stereochemistry between the two fluoroketals, since the S epimer 32 showed splitting of the lea-methyl by an adjacent fluorine,

(26) (a) For a review, see: Mukaiyama, T. A&u. &en. 1979,18,707. (b) For use of FMPT in thioester synthesis, see: Watanabe, Y.; Shoda, S.; Mukaiyama, T. Chem. Lett. 1976, 741.

(27) The participation of fluoride ion in the absence of other nucleo- philes is the basis of a method for the synthesis of acyl fluorides using FMPT: Mukaiyama, T.; Tanaka, T. Chern. Lett. 1976, 303.

2320 J. Org. Chem., Vol. 51, No. 12, 1986 Kertesz and Marx

Scheme IV. Reactions of l7a-Acetoxy Acids w i t h FMPT

F

Y

13 - 0 J-JJ whereas the R epimer 33 did not.28 The 160-methyl-17- acetoxy acid 13 was similarly converted into epimeric fluoroketals 34 and 35, but the relative stereochemistry of this pair could not be assigned by NMR, presumably since the 16@-methyl was too far from the ketal fluorine to show a coupling.

The standard FMPT/NaSCH3 conditions were applied to 16fl-methyl-17-hydroxy acid 9 in the hope that the lower reaction temperature would spare the 9,ll-dichloride system. The dichloride system did indeed survive, but in this case the major product isolated in 34% yield was that of dehydration, the A%arboxylic acid 36 (Scheme V). The 9,ll-fluorohydrin lea-methyl acid 6 behaved quite differently when used in the same reaction sequence. In this case, the reaction mixture yielded 30% of normal thiol ester 37,18% of the 18n0r-A'~*'~-17-methyl rearrangement product 38, and 12% of an unstable compound to which the 13,17-lactone-l7@-methyl structure 39 has been assigned. This labile material absorbed at 1810 cm-' in the infrared, and was transformed quantitatively into 38 with apparent loss of carbon dioxide while drying at 100 "C. The properties of 39 are compatible with the assigned structure as well as its possible role as an intermediate in the formation of olefin 38. Mercaptide anion was not essential to the formation of these rearrangement products, and reaction of 6 with FMPT and TEA alone gave 38 (16%) and 39 (11%) along with some starting acid (7%) as the only isolable products. The 17-methyl olefin substrate exemplified by 38 is familiar among products of the Miescher-Kligi rearrangementm of androstane 17-alcohols and tosylates,30 and the formation of @-lactone 39 appears related to a known rearrangement of steroid 17-spiro ox- e t a n ~ n e s . ~ l t ~ ~

(28) The uae of long range coupling effects to determine the stereochemistry of fluoride substituents is precedented by studies of the interaction between the steroid 6B-fluorine and the 19-angular methyl group: Bhacca, N. S.; Williams, K. H. Applications of NMR Spectros- copy in Organic Chemistry; Holden-Day: San Francisco, 1964; pp 123-134, and references therein.

(29) (a) Kigi, H.; Miescher, K. Helu. Chim. Acta 1939, 22, 683. (b) Miescher, K.; K&i, H. Ibid 1949, 32, 761.

(30) For reviews, see: (a) Wendler, N. L. In Molecular Rearrange- ments Part 2; de Mayo, P., Ed.; Wiley-Interscience: New York, 1964; pp 1020-1026. (b) Kirk, D. N.; Hartshorn, M. P. Steroid Reaction Mecha- nisms; Elsevier: New York, 1968; pp 269-272.

- 35: X=F,Y=CE3

Stereochemistry at (2-16 was suspected of influencing the choice of reaction path between products of 18-methyl migration and 16,17-olefin formation. Therefore, two additional 160-methyletianic acids were subjected to the FMPT/TEA reaction conditions. The 9,ll-fluorohydrin 7 gave 48% of 17-methyl-A13J7-olefin 40 and 8% of the labile @-lactone 41, in results which parallel those for lea-methyl fluorohydrin 6. Addition of NaSCH, to the reaction mixture did not change yields or give thiol ester products. The A9J1-acid 8 behaved much the same as 9,ll-dichloride 9 and gave 31% of dehydration product acid 42. A 15% yield of corresponding thiol ester 43 accompanied 42 when NaSCH, was added to the reaction mixture, whereas the Als-acyl fluoride 44 was isolatedn in place of 43 when thiolate was not used. It was not possible to detect any overlap of product types in the reported reactions: each 17-hydroxyetianic acid appeared to give either 18-methyl migration or A16-products exclusively.

Thus, the difference in reactions of 17-hydroxy acids with FMPT was associated with the 9,ll-substitution pattern rather than (2-16 stereochemistry. Both the 16a and the 160-methyl fluorohydrins 6 and 7 gave 18-methyl migration products, whereas dichloride 5 and 9,ll-olefin 4 experienced 17-alcohol dehydration?, The influence of the C-ring substitution pattern on the choice of reaction path was most likely due to its effect on overall molecular conf~rmat ion ,~~ probably in determining the nature and strain energy of the actual intermediate species. The

(31) (a) Hen, J. E.; Fried, J.; Grabowich, P.; Sabo, E. F. J. Am. Chem. SOC. 1956,78,4812. (b) Hirschmann, R.; Bailey, G. A.; Poos, G. I.; Walker, R.; Chemerda, J. M. Ibid 1966, 78,4814.

(32) Rearrangement of 17,21-oxetanones to the isomeric l7g-methyl- 13ql7a-keto oxides (see ref 31) is an example of 18-methyl migration with participation of a neighboring oxygen functionality (see ref 30a; also ref 30b, p 369).

(33) A similar divergence in reactivity reflecting C-ring substituent effects has been observed in acid-catalyzed reactions of 17a-hydroxy steroid 3,20-diaemicarbnes, wherein 11-ketones gave A1s-produda (see ref 34) and 11-alcohols underwent 18-methyl migration (see ref 35).

(34) Slates, H. L.; Wendler, N. L. J. Org. Chem. 1957,22, 498. (35) Taub, D.; Hoffsommer, R. D.; Wendler, N. L. J. Org. Chem. 1964,

29, 3486. (36) The potential of C-ring substitution to markedly alter the pre-

ferred conformation of the steroid side chain was illustrated in a recent X-ray crystallographic study of a series of pregnanes: Duax, W. L.; Griffin, J. F.; Rohrer, D. C. J . Am. Chem. SOC. 1981, 103, 6705.

Thiol Esters from Steroid 17P-Carboxylic Acids J. Org. Chem., Vol. 51, No. 12, 1986 2321

H,C, OOTs

Figure 2.

(iii)

\ -co*

Scheme V. Reactions of 17a-OH Acids with FMPT and NaSCH,

tendency of the resulting intermediate to rearrange would then depend on the degree of transcoplanarity between the 18-methyl and 17-leaving groups attained therein.30 Both 17-oxo adduct (ii) and a-lactone3’ (iii) are potentially ac-

cessible from base-catalyzed transformations of initial pyridinium adduct (i), and either could serve as a transient intermediate in the subsequent reactions (Figure 2).

Activation with Carbonyldiimidazole. The use of carbonyldiimidazole38 (CDI) as an activating agent pro-

(37) Chapman, 0. L.; Wotjkowski, P. w.; Adam, w.; Rodriguez, 0.; Ruckthchel, R. J. Am. Chem. SOC. 1972,94,1365, and references therein. (38) Gaia, H. J. Angew. Chem. 1977, 16, 244.

2322 J . Org. Chem., Vol. 51, No. 12, 1986

Scheme VI. Thiol Esters of 17a-OH Acids via CDI

COOH

Kertesz and Marx

i

COOH ONc S C H ,

7 1 - I F 1 46 w - - 0'

4 7 -

i vided a solution to problems of thiol ester formation in the presence of both basesensitive C-ring substituents and free 17-hydroxy groups. The intermediates formed in reactions of CDI with free 17-hydroxy acids comprised the only species encountered in the present work that was reactive toward free mercaptans. Thus, treatment of 16a- methyl-17-hydroxy acid 6 with CDI followed by addition of ethanethiol gave ethylthio ester 45 (40%), and 16p- methyl methylthio ester 46 (61 %) was obtained by bubbling methyl mercaptan through the reaction mixture of CDI with 17-hydroxy acid 7 in DMF (Scheme VI). Syntheses of 9,11-dichloro-17-acetate methylthio ester 48 and propionate 49 were then accomplished by a similar preparation of 17-hydroxy thiol ester 47 from 9, followed by dimethylaminopyridine (DMAP) catalyzed esterifica- tions of the hindered 17-alcohol in mixtures of acetic or propionic anhydride with TEA at 80 "C. No reaction was observed when the 17-acetoxy acids 11 and 13 and the 16,17-acetonide 10 were treated with CDI followed by either mercaptan or mercaptide anion, presumably because reactive intermediates were not formed. The ability of CDI to activate 17-hydroxy acids towards thiol ester formation is remarkable in contrast to the failures with 17-derivatized substrates, and could result from steric factors alone or in combination with actual participation by the free 17- alcohol. In view of the ease with which 17-spirocyclic derivatives of the 17-hydroxy acids are formed, it seems reasonable to propose that the initial adduct (i) could give

the cyclic carbonate (ii) as the actual intermediate in these cases,39 instead of an a~ylimidazolide.~~

Many of the etianic acid thiol esters prepared in this work have proven to be extremely potent topical antiinflammatory agents, as measured in the rat40 and in human vasoconstrictofi' assays. A noteworthy separation between topical and systemic corticoid activity was also achieved, since most of the compounds were relatively inactive in the rat thymus involution and carrageenan-paw assays.42 Thus, the 16a-methyl-17-propionyloxy methylthio ester 21 possessed topical activity in the order of fluocinolone acetonide (1) and the 16P-methyl 17-acetate 24 was approximately 1.3 times as potent, whereas both compounds demonstrated systemic activities of less than 5 times hy- droc~rt isone.~~ A high ratio of topical to systemic activity such as that displayed by thiol esters 21 and 24 is of potential therapeutic value.2b All examples of 17-spiro acyl ortho esters and fluoro ketals were virtually inactive in both types of assay.

Experimental Section General Methods. Melting points were obtained by using a

Fisher-Johns apparatus and are uncorrected. Analytical TLC was

(39) We thank Dr. Arthur Kluge for this suggestion. (40) Tonelli, G.; Thibault, L.; Ringler, I. Endocrinology 1965, 77,625. (41) Place, V. A.; Velasquez, J. G.; Burdick, K. H. Arch. Dermatol.

1970, 101, 531. (42) Potencies were determined in intact rats using a combination of

a two-day thymolytic assay (see ref 43) and a paw edema test (see ref 44). (43) Dorfman, R. I.; Dorman, A. S.; Agnello, E. J.; Figdor, S. K.;

Lauback, G. D. Acta Endocrinol. (Copenhagen) 1961, 37, 343. (44) Winter, C. A.; Risley, E. A.; Nass, G. W. Proc. SOC. Erp. Biol.

Med. 1962, I l l , 544. (45) The topical activity of hydrocortisone is approximately 0.001

times that of fluocinolone acetonide 1 when measured in our vasoconstrictor assays (see ref 41).

Thiol Esters from Steroid 17P-Carboxylic Acids

performed with Analtech 2.5 cm X 10 cm X 0.25 mm silica GF plates, and preparative TLC was done on Analtech silica GF 20 cm X 40 cm X 1 mm plates or with a Harrison Chromatotron centrifugal chromatography apparatus using 1 mm and 2 mm thick silica PF plates. Infrared spectra were recorded with a Perkin- Elmer Model 137 spectrophotometer, and frequencies are quoted to the nearest 5 cm-'. Ultraviolet spectra were taken in methanol on a Cary 14 spectrophotometer. Proton NMR spectra were determined on Varian HA-100 (100 MHz), Bruker WH-90 (90 MHz), or Bruker WM-300 (300 MHz) instruments as solutions in hexadeuteriodimethyl sulfoxide (Me2SO) or deuteriochloroform (CDC13) containing tetramethylsilane as an internal reference, and carbon-13 spectra were recorded on the Bruker WM-300 (75.5 MHz). Electron impact mass spectral data were obtained with Finnegan MAT CH-7 and 1 2 2 4 direct inlet instruments at 70 eV, chemical ionization spectra were measured on the MAT 1 2 2 3 at 90 eV, and high-resolution mass spectrometric measurements were made on a Finnigan MAT-311A. Elemental analyses were performed by the Syntex Analytical Services Group. All reactions were performed in heat-dried glassware under an atmosphere of nitrogen, unless otherwise noted, and sensitive reagents were introduced through septa by syringe.

Materials. Fluocinolone acetonide (1 , 6a,9a.-difluoro- ll/3,l6a,l7a,21-tetrahydroxypregna-1,4-diene-3,2O-dione 16,17- acetonide), flumethasone (2, 6a,9a-difluoro-ll/3,17a,21-trihydroxy-16a-methylpregna-1,4-diene-3,20-dione), and betamethasone (3, 9a-fluoro-11/3,17a,21-trihydroxy-16/3-methy1- pregna-1,4-diene-3,2O-dione) are commercially available. The A9%orticoid 4 was kindly supplied by the Upjohn Company in the form of the 17,21-diacetate (17,21-dihydroxy-6a-fluoro-l6/3- methylpregna-1,4,9(1l)-triene-3,20-dione 17,21-dia~etate) .~~ Chlorophosphate reagents, CDI, and DMAP were purchased from Aldrich; FMPT (prepared from 2-fl~oropyridine,~ Aldrich) is now commercially available. Tetrahydrofuran (THF) was purified by distillation from Na benzophenone. Standard solutions of NaSCH3 (usually 0.83 N) were prepared by bubbling methyl mercaptan through a slurry of NaH (50% in oil, usually 2.0 g) in DMF (50 mL) until all solids dissolved, and solutions of other mercaptide salts were made similarly from liquid mercaptans and NaH in DMF or THF.

17,2 1 -Dihydroxy-6a- fl uoro- 168-met hylpregna- 1,4,9( 1 1 ) - triene-3,20-dione (4). To a solution of 17,21-dihydroxy-6a- fluoro-16/3-methylpregna-1,4,9(1l)-triene-3,20-dione 17,21-di- acetate46 (1 g, 2.18 mmol) in MeOH (40 mL) was added K2CO3 (0.5 g) in water (5 mL), and the mixture was stirred at 20 "C for 1 h. The solution was acidified with 6 N HC1 and diluted with ice water, after which the precipitate was filtered, washed, and dried to give 745 mg (91%) of crystalline 4, mp 198-200 "C. A sample was recrystallized from acetone-hexane for analysis: mp 202-204 "C; UV 238 nm (t 16400); IR (KBr) 1720 ( 2 0 - M ) , 1665 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.73 (5, 3 H, 18- CH3), 1.03 (d, 2 H, J = 6.4 Hz, 16-CH3), 1.37 (s, 3 H, 19-CH3),

5.60 (d, 1 H, J = 4.6 Hz, 11-H), 5.71 (br dm, 1 H, J = 45 Hz, 6-H),

(d, 1 H, J = 10 Hz, 1-H); MS, m / e 374 (M+). Anal. Calcd for C22H2704F: C, 70.57; H, 7.27. Found: C, 70.56; H, 7.17.

9a,l l@-Dichloro- 17a,2 1-dihydroxy-6a-fluoro- 1 6B-met hylpregna-1,4-diene-3,2O-dione (5). A solution of 17,21-dihydroxy-6a-fluoro-16/3-methylpregna-1,4,9( 1 l)-triene-3,20-dione 17,21-diacetate4 (1 g, 2.18 mmol) in a mixture of CC14 (20 mL) and pyridine47 (3 mL) was purged with nitrogen for 5 min, then treated with a gentle stream of chlorine for 15 min, and flushed with nitrogen again for 5 min. The resulting mixture was diluted with EtOAc, washed with dilute HCl, water, and brine, dried (Na2S04), and evaporated to dryness. The residue was recrystallized twice from acetone-hexane, giving 840 mg of the intermediate 9,11-dichloro-17,21-diacetate, which was then slurried in a mixture of MeOH (40 mL) and a solution of K2C03 (300 mg) in water (4 mL) for 30 min. Ice water (30 mL) was added and

4.18 (d, 1 H, J = 19.5 Hz, 21-H), 4.38 (d, 1 H, J = 19.5 Hz, 21-H),

6.07 (9, 1 H, 4-H), 6.20 (dd, 1 H, 51 = 10 Hz, 52 = 2 Hz, 2-H), 7.39

J. Org. Chem., Vol. 51, No. 12, 1986 2323

the resulting precipitate was filtered, washed, and dried to give 590 mg (61%) of crystalline 5: mp 215-217 "C; UV 236 nm (e 15000); IR (KBr) 1710 (20-C=0), 1670 (3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 1.07 (s, 3 H, 18-CH3), 1.08 (d, 3 H, J = 6 Hz, 16-CH3), 1.71 (s, 3 H, 19-CH3), 4.15 (d, 1 H, J = 19 Hz, 21-H), 4.39 (d, 1 H, J = 19 Hz, 21-H), 4.96 (br m, 1 H, 11-H), 5.58 (br dm, J = 48 Hz, 6-H), 6.10 (s, 1 H, 4-H), 6.30 (dd, 1 H, J1 = 10 Hz, J2 = 2 Hz, 2-H), 7.24 (d, 1 H, J = 11 Hz, 1-H); MS, m / e 426, 428 (M+ - H20). Anal. Calcd for C,HnO4Cl2F: C, 59.33; H, 6.11. Found: C, 59.32; H, 5.98.

Oxidations with Periodic Acid. 6a,9a-Difluoro-l1/3,17a- dihydroxy-16a-methyl-3-oxoandrosta-1,4-diene-l7~- carboxylic Acid (6). To a slurry of (2) (2 g, 5.11 mmol) in MeOH (40 mL) was added a solution of periodic acid (1.7 g, 7.46 mmol) in water (40 mL), and the mixture was stirred open to the air at 20 O C for 16 h. The volume was reduced to 60 mL by evaporation, ice water (80 mL) was added, and the precipitate was filtered, washed, and dried, giving 1.88 g (94.5%) of crystalline 6, mp 285-289 "C. A sample was recrystallized from acetone-hexane for analysis:48 nip 303-304 "C dec; UV 238 nm (e 16300); IR (KBr) 1700 (COOH), 1660 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.87 (d, 3 H, J = 9 Hz, 16-CH3), 1.00 (8, 3 H, 18-CH3), 1.49 (s, 3 H, 19-CH3), 4.15 (br d, 1 H, J = 10 Hz, 11-H), 5.64 (br dm, 1

J2 = 2 Hz, 2-H), 7.26 (d, 1 H, J = 9.4 Hz, 1-H); MS, m / e 396 (M'). Anal, Calcd for C21H2605F2: C, 63.63; H, 6.61. Found: c, 63.69; H, 6.72.

17-Hydroxyetianic acids 7-9 were prepared by the same procedure.

11~,17a-Dihydroxy-9a-fluoro-16~-methy1-3-oxoandrosta- 1,4-diene-17B-carboxylic acid (7) was prepared from 3 in 97% yield: mp 247-251 "C. A sample was purified by dissolving in dilute Na2C03, washing with EtOAc, acidifying the aqueous phase with dilute HCl, and cooling to ice. The precipitate was collected, washed, dried, and recrystallized from EtOAc:" mp 254-256 "C (lit.lo mp 256-258 "C); UV 239 nm (t 21 450); IR (KBr) 1750,1720 (COOH), 1665 ( 3 4 4 ) cm-'; 'H NMR (300 MHz, Me2SO) 6 1.07 (s, 3 H, 18-CH3), 1.14 (d, 3 H, J = 9 Hz, 16-CH3), 1.50 (s, 3 H, 19-CH3), 4.14 (br d, 1 H, J = 12 Hz, 11-H), 6.01 (s, 1 H, 4-H), 6.21

1-H); MS, m / e 378 (M'). Anal. Calcd for C21H2705F: C, 66.65; H, 7.19. Fodnd C, 66.40; H, 7.47. 6a-Fluoro-17a-hydroxy-16~-methyl-3-oxoandrosta-1,4,9-

(ll)-triene-l7~-carboxylic acid (8) was prepared from 4 in 92% yield: mp 200-201 "C dec. Analytical sample from acetone- hexane: mp 197-199 "C dec; W 238 nm (e 16200); IR (KBr) 1740, 1710 (COOH), 1655 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.86 (s, 3 H, 18-CH3), 1.13 (d, 3 H, J = 6.6 Hz, 16-CH3), 1.38 (s, 3 H, lg-CH,), 5.61 (d, 1 H, J = 5.2 Hz, 11-H), 5.72 (br dm, 1

J2 = 1.4 Hz, 2-H), 7.40 (d, 1 H, J = 10 Hz, 1-H); MS, m / e 360 (M'). Anal. Calcd for CzlHz504F: C, 69.98; H, 6.99. Found: C, 69.99; H, 7.24. 9a,l1~-Dichloro-6j3-fluoro-l7a-hydroxy- 168-methyl-3-

oxoandrosta-1,4-diene- 17@-carboxylic acid (9) was prepared from 5 in 91% yield: mp 221-223 "C dec; UV 236 hm (c 15200); IR (KBr) 1750,1710 (COOH), 1665 (3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 1.14 (s, 3 H, 18-CH3), 1.17 (d, 3 H, J = 8 Hz, 16-CH,), 1.69 (s, 3 H, 19-CH3), 4.96 (br m, 1 H, 11-H); MS, m / e 430,432 (M+). Anal. Calcd for C21H2504C12F: C, 58.48; H, 5.84. Found: C, 58.59; H, 5.78.

Base-Catalyzed Air Oxidation. Ga,Sa-Difluoro- 11~,16a,17a-trihydroxy-3-oxoandrosta- 1,4-diene-l78- carboxylic Acid 16,17-Acetonide (10). A stream of air was bubbled under the surface of a slurry of 1 (0.5 g, 1.10 mmol) in MeOH (25 mL) containing K&03 (0.5 g) for 10 min, after which the air flow was stopped and the mixture was stirred open to the air for another 16 h at 20 "C. The volume was reduced to 10 mL by evaporation, after which EtOAc (80 mL) was added and the mixture was extracted with water. The aqueous layer was acidified

H, J = 48 Hz, 6 H), 6.10 (s, 1 H, 4-H), 6.27 (dd, 1 H, 51 = 10 Hz,

(dd, 1 H, 51 = 10 Hz, J2 = 2 Hz, 2-H), 7.29 (d, 1 H, J = 10 Hz,

H, J = 48 Hz, 6-H), 6.07 (s, 1 H, 4-H), 6.20 (dd, 1 H, 51 = 10 Hz,

(46) For preparation, see: Ayer, D. E.; Schlagel, C. A.; Flynn, G. L. Ger. Patent 2 308 731, 1973; Chem. Abstr. 1973, 79, 146739a.

(47) For a prior use of pyridine as a solvent for chlorination of steroids, see: Cutler, F. A., Jr.; Mandell, L.; Shew, D.; Chemerda. J. M. J. Org. Chem. 1959,24, 1621.

(48) Full characterization was performed because the literature

(49) Anner, G.; Meystre, Ch., US. Patent 3636010, 1972. (50) This compound has been reported, but spectral data were incom-

melting point of 333 OC (see ref 49) was unattainable.

plete: cf. ref 10.

christinehurd

Highlight

christinehurd

Highlight

2324 J. Org. Chem., Vol. 51, No. 12, 1986

with dilute HC1 and reextracted with EtOAc, the resulting organic layer being washed with water and brine, dried (Na2S04), and evaporated to dryness. Recrystallization from acetone-hexane gave 370 mg (75%) of 10: mp 301-303 "C dec; UV 237 nm (c 15900); IR (KBr) 1725 (COOH), 1665 (3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 0.93 (s, 3 H, 18-CH,), 1.14 and 1.29 (2 s, 6 H, C(CH,),), 1.48 (s, 3 H, 19-CH3), 4.15 (br d, 1 H, J = 10 Hz, 11-H), 4.96 (m, 1 H, 6-H); MS, m/e 438 (M'). Anal. Calcd for C2,H,,F2O,: C, 63.00; H, 6.44. Found: C, 63.13; H, 6.24. 17a-Acetoxy-6a,9a-difluoro-11@-hydroxy-16a-methy1-3-

oxoandrosta-1,4-diene-l7@-carboxylic Acid (1 1). Acetyl chloride (2.5 mL) was added to an ice-cooled solution of 6 (1.5 g, 3.78 mmol) in CHzClz (50 mL) and TEA (7.5 mL), and the mixture was stirred a t 0 "C for 30 min, after which TLC (1% HOAc-35% acetone-hexane) indicated that starting material had been replaced by a mixture of 11 with a less polar material assumed to be the mixed anhydride of acetic acid with 11. Diiso- propylamine (7.5 mL) was added and the mixture was stirred for 2 h at 20 "C and then evaporated to dryness. The residue was dissolved in EtOAc and the solution washed with dilute HC1, water, and brine, dried (Na2S04), and evaporated. The residue was triturated with hot CH2C12, affording 1.15 g (69%) of 11. The analytical sample was recrystallized 4 times from acetonehexme: mp 180-182 "C dec; UV 238 nm (t 15 800); IR (KBr) 1735 (br, COOH, ester), 1665 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.86 (d, 3 H, J = 7 Hz, 16-CH3), 1.02 (s, 3 H, 18-CH,), 1.49 (s, 3 H, lg-CH,), 2.01 (s, 3 H, COCH,); MS, m/e 418 (M' - HF). Anal. Calcd for Cz3Hzs06F2: C, 63.00; H, 6.44. Found: C, 62.87; H, 6.28.

Ga,Sa-Difluoro- 1 l@-hydroxy-17a-(propionyloxy)- 16a- methyl-3-oxoandrosta- 1,4-diene-l7@-carboxylic acid (12) was prepared from 6 by using propionyl chloride in the above procedure. Recrystallization from acetone-hexane gave 94% of 12, mp 217-218 "C; W 238 nm ( E 16825); IR (KBr) 1730 (ester), 1700 (COOH), 1665 ( 3 4 4 ) cm-l; 'H NMR (100 MHz Me2SO) 6 0.83 (d, 3 H, J = 7 Hz, 16-CH,), 0.98 (t, 3 H, J = 7 Hz, CH,CH,), 0.99 (s, 3 H, 18-CH3), 1.47 (9, 3 H, 19-CH3), 2.26 (q, 2 H, J = 7 Hz, COCH2); MS, m / e 350 (M+ - COOH, COC,H,). Anal. Calcd for C24H300BF2: C, 63.71; H, 6.68. Found: C, 63.74; H, 6.73.

17a-Acetoxy-9a-fluoro-1 1@-hydroxy-16@-methy1-3-oxoandrosta-l,4-diene-l7@-carboxylic Acid (13). Acetyl chloride (6 mL) was added to a solution of 7 (4.60 g, 12.2 mmol) in CH2Cl2 (200 mL) and TEA (4 mL), and the mixture was stirred at 0 "C for 30 min. Diisopropylamine (25 mL) was added and the solution was heated at reflux for 1 h, after which the procedure was completed as described for 11. Crystallization from acetone- hexane gave 4.76 g (93%) of 13, mp 196-198 "C. Analytical sample from acetone-hexane:" mp 218-220 "C (lit.lo 212-214 "C); UV 239 nm (e 17900); IR (KBr) 1735 (br, COOH, ester), 1655 ( 3 - M ) cm-'; 'H NMR (300 MHz, Me,SO) 6 0.98 (s, 3 H, 18-CH3), 1.31 (d, 3 H, J = 7 Hz, 16-CH3), 1.50 (s, 3 H, 19-CH3), 1.94 (s, 3 H, COCH,); MS, m / e 361 (MH' - HOAc). Anal. Calcd for C23H290BF: C, 65.70; H, 6.95. Found: C, 65.73; H, 7.01.

Sa-Fluoro-1 l@-hydroxy-l7a-(propionyloxy)-16@-methyl-3- oxoandrosta-1,4-diene-17@-carboxylic Acid (14) and 9a- Fluoro-1 18- hydroxy- 17a-(propionyloxy)-l6@-methyl-3-oxoandrosta-1,4-diene-l7@-carboxylic Propionic Anhydride (16). Propionate 14 was prepared in 84% yield from 7 by using propionyl chloride in the above procedure. The analytical sample was recrystallized three times from acetone-hexane:" mp 182-183 "C dec (lit. 'O mp 188-190 "C; W 239 nm (e 14000); IR (KBr) 1735 (br, COOH, ester), 1655 (3-C=0) cm-'; 'H NMR (100 MHz, Me,SO) 6 1.06 (s, 3 H, 18-CH3), 1.06 (t, 3 H, J = 7 Hz, CH2CH3), 1.39 (d, 3 H, J = 7 Hz, 16-CH3), 1.53 (s, 3 H, 19-CH3), 2.25 (q, 2 H, J = 7 Hz, COCH,); MS, m / e 435 (MH'). Anal. Calcd for C,H,,06F C, 66.34; H, 7.19. Found C, 66.15; H, 7.08. A second reaction using 75 mg (0.20 mmol) of 7 was worked up directly, without the diisopropylamine step, affording 80 mg (72%) of 16, mp 170-173 "C. The analytical sample was recrystallized from acet~ne-hexane:~~ mp 177-178 "C dec (lit.'O 180-182 "C); UV 239 nm (c 16240); IR (KBr) 1815 (anhydride), 1730 (br, anhydride ester), 1660 (3-C=0) cm-'; 'H NMR (100 MHz, CDC1,) 6 1.10, 1.14 (2 t, 6 H, J's = 7 Hz, CH,CHis), 1.11 (s, 3 H, 18-CH3), 1.42 (d, 3 H, J = 7 Hz, 16-CH3), 1.55 (s, 3 H, 19-CH3), 2.30, 2.43 (2 q, 4-H, S s = 7 Hz, COCH,'s); MS, m / e 470 (M' - HF). Anal. Calcd for CzzH,607F: C, 66.11; H, 7.19. Found: C, 66.09; H, 7.16.

Kertesz and Marx

9a,ll@-Dichloro-6a-fluoro-l7a-(propionyloxy)-l6@- methyl-3-oxoandrosta-l,4-diene-l7@-carboxylic Acid (15). Treatment of 9 (352 mg, 0.82 mmol) with propionyl chloride and TEA followed by diisopropylamine gave 229 mg (58%) of 15 (from acetone-hexane), mp 178-180 "C dec; UV 236 nm (6 14400); IR (KBr) 1740 (ester), 1710 (COOH), 1665 (3-C=0) cm-'; 'H NMR (300 MHz, Me,SO) 6 1.00 (t, 3 H, J = 7.5 Hz, CH,CH3), 1.05 (s, 3 H, 18CH3), 1.35 (d, 3 H, J = 7 Hz, 16-CHJ, 1.72 (s, 3 H, 19-CHJ,

MS, m / e 487 (MH'). Anal. Calcd for Cz,HmO5Cl2: C, 59.14; H, 6.00. Found: C, 59.11; H, 6.04.

General Methods for Synthesis of Thiol Esters from Acids 10-14. Preparation of 17@-[(Methylthio)carbonyl]-6a,9a- difluoro- 11@,16a,17a-trihydroxyandrosta- l,4-dien-3-one 16,17-Acetonide (17). Method A. A slurry of 10 (351 mg, 0.80 mmol) in THF (10 mL) containing TEA (0.20 mL, 1.44 mm), was stirred at 20 "C for 30 min, during which time the acid dissolved and a precipitate assumed to be the triethylammonium salt was formed. Diethyl chlorophosphate (0.17 mL, 1.20 mmol) was added and stirring was continued, giving a new precipitate of triethylammonium chloride (Et3N-HC1). After 2 h, TLC (3% MeOH-CH,Cl,) of a sample of reaction mixture worked up between EtOAc and dilute aqueous HC1 showed that conversion of 11 to a neutral intermediate was complete. Filtration into a fresh flask gave 118 mg (0.85 mmol) of pure Et3N.HC1 (elemental analysis) which was discarded. An 0.83 N solution of NaSCH, in DMF (2.2 mL, 1.83 mmol) was added to the filtrate, and the mixture was stirred at 20 "C for 16 h, after which it was diluted with EtOAc and extracted with dilute aqueous Na2C03. The organic layer was washed, dried (Na2S04), and evaporated to dryness, giving the crude thiol ester as a very poorly soluble solid. Recrystallization from 1:2 MeOH/CH2C12 by boiling off the CHzC12 and cooling in ice gave 255 mg (68%) of pure 17, mp >300 "C; UV 238 nm (c 20415); IR (KBr) 1665 (br, 20-C=0,3-C=0) cm-'; 'H NMR (100 MHz, MezSO) 6 0.83 (s, 3 H, 18-CH3), 1.13,

MS, m / e 453 (M+ - CH3). Anal. Calcd for CaH,05FS: C, 61.52; H, 6.45; S, 6.84. Found: C, 61.40; H, 6.56; S, 6.55. Acidification of the Na2C03 extract, extraction with EtOAc, and recrystallization from acetone-hexane gave 64 mg (18%) of pure starting acid 10.

Method B. Diphenyl chlorophosphate (116 pL, 0.5 mmol) was added to a solution of 10 (175 mg, 0.40 mmol) in THF (6 mL) and TEA (100 pL, 0.72 mmol), and the mixture was heated at 55 "C for 2 h, after which the TLC of a worked-up sample showed that formation of a neutral intermediate was complete. The mixture was cooled in ice, filtered to remove the Et,N.HCl (52 mg, 0.38 mmol), and 0.83 N NaSCH3/DMF (1.0 mL, 0.83 mm) was added. The mixture was stirred at 20 "C for 1 h and worked up as described in method A, affording (after recrystallization from MeOH-CH2C12) 141 mg (75%) of pure 17, mp >300 "C.

Method C. To a slurry of FMPT (90 mg, 0.32 mmol) in CH2C12 (3 mL) at -15 "C was added a separately prepared solution of 10 (110 mg, 0.25 mmol) in a mixture of CHzClz (3 mL) and TEA (58 pL, 0.42 mmol). The conversion of 10 to a nonpolar intermediate was complete in 20 min (TLC, 1% HOAc-40% EtOAe-hexane), after which a solution of NaSCH, (0.83 N in DMF, 1.0 mL, 0.83 mmol) was added and the mixture was maintained at -15 "C in a freezer for 16 h. Standard acid-base workup and recrystallization from MeOH then afforded 109 mg (93%) of pure 17 in two crops, mp >300 "C.

17&[ (Ethylthio)carbonyl]-6a,9a-difluoro-11@,16a,17a-trihydroxyandrosta-1,4-dien-3-one 16,17-Acetonide (18). Pre- pared from 10 (350 mg, 0.80 mmol) according to method A, using NaSC2H, [ from ethanethiol (0.20 mL) and NaH (50% in oil, 68 mg, 1.42 mmol) in THF (3 mL)]. Recrystallization from acetone-hexane gave 240 mg (65%) of 18 in two crops: mp >300 "C; UV 239 nm (6 20 145); IR (KBr) 1665 (br, 20-C=0,3-C=0); 'H NMR (100 MHz, Me,SO) 6 0.85 (s, 3 H, 18-CH3), 1.13, 1.32

H, J = 7, SCH,); MS, m / e 467 (M' - CH,). Anal. Calcd for C25H3205F2S: C, 62.22; H, 6.68. Found: C, 62.11; H, 6.75.

l7a-Acetoxy-6a,9a-difluoro-l l@-hydroxy-17&[ (methylthio)carbonyl]-l6a-methylandrosta-1,4-dien-3-one (19) and Ga,Sa-Difluoro- 11@,17a-dihydroxy- l6a-methyl-3-oxoandrosta-1,4-diene-l7j3-~arboxylic Acid 17,20-(1'-Methyl- 1'-(methy1thio))methylene Ketal (23). A solution of the in-

2.27 (9, 2 H, J = 7.5 Hz, COCH,), 5.06 (d, 1 H, J = 2.5 Hz, 11-H);

1.33 (2 8, 6 H, C(CH,)z), 1.46 ( ~ , 3 H, Ig-CH,), 2.29 (s, 3 H, SCH,);

(2 S, 6 H, C(CH3)2, 1.17 (t, 3 H, J = 7 Hz, CHPCH,), 2.83 (4, 2

christinehurd

Highlight

Thiol Esters from Steroid 17P-Carboxylic Acids

termediate prepared from 11 (420 mg, 0.96 mmol) according to method A was treated with 0.83 N NaSCH,/DMF (4 mL, 3.32 mmol) and stirred at 20 "C for 16 h. Workup gave 200 mg (48%, after recrystallization) of pure starting acid 11 and a neutral fraction which was recrystallized from acetone-hexane, affording 58 mg (13%) of thiol ester 19. Analytical sample of 19 (from acetone-hexane): mp >300 "C; UV 238 nm ( e 20100); IR (KBr) 1750 (ester), 1670 (20-C=0,3-C=0) cm-'; 'H NMR (300 MHz, Me,SO) 6 0.91 (d, 3 H, J = 7 Hz, 16-CH3), 0.97 (s, 3 H, 18-CH3),

MS, m/e 469 (MH'). Anal. Calcd for C2.,H3005FZS: C, 61.52; H, 6.45. Found: C, 61.63; H, 6.38. The neutral mother liquor material consisted of a mixture of 19 with a slightly less polar material which was probably the spiro ortho ester 23. In a separate experiment, a reaction mixture prepared from 11 (400 mg, 0.91 mmol), diethyl chlorophosphate (0.25 mL, 1.73 mmol) and TEA (0.20 mL, 1.43 mmol) in THF (10 mL) was treated with 0.4 mL increments of 0.83 N NaSCH,/DMF at 15-min intervals until the TLC (3% MeOHCH,Cl,, developed twice) of worked-up samples showed complete replacement of phosphate intermediate (Rf 0.34) by thiol ester 19 (Rf 0.47), ortho ester 23 (Rf 0.55), and acid 11 (Rf 0.0). A total of 2.8 mL (2.3 mmol) of NaSCH, solution was required. Acid-base workup gave 136 mg (34%) of acid 11, and crystallization of the neutral residue from acetone-hexane gave mixtures 19 and 23 consisting of a first crop (113 mg) enriched in 19 and a second crop (98 mg) enriched in 23. The mixtures were separated by preparative TLC (30% EtOAc-hexane, eluted twice), and recrystallizations (acetone-hexane) then gave 40 mg (9%) of 19 and 52 mg (12%) of 23: mp 201-203 "C; UV 237 nm ( t 16670); IR (CHC1,) 1790 (20-C=0), 1670 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.95 (d, 3 H, J = 7 Hz, 16-CH3), 1.26 (s, 3 H, 18-CH3), 1.50 (s, 3 H, 19-CH3), 1.92 (8, 3 H, l'-CH3), 2.22 (s,3 H, SCH,); MS, m/e 469 (MH'). Anal. Calcd for CuH&$zS: C, 61.52; H, 6.45. Found C, 61.50; H. 6.37. Addition of a &fold excess of NaSCH, to a solution of 23 in THF gave complete conversion to 19 in 2 h at 20 "C.

Thiol ester 19 was also prepared as follows: thionyl chloride (40 pL, 0.56 mmol) was added to solution of 11 (70 mg, 0.16 mmol) in THF ( 5 mL) containing TEA (0.10 mL, 0.72 mmol) at 0 "C, and the mixture was filtered into a fresh flask after 10 min. Addition of 0.83 N NaSCH,/DMF (0.8 mL, 0.66 mmol) to the filtrate and workup after 1 h at 20 "C gave 24 mg (32%) of 19, mp >300 "C, and 11 mg (16%) of 11.

17a-Acetoxy-6a,9a-difluoro-l7&[ (ethy1thio)carbonyll- 11~-hydroxy-l6a-methylandrosta-l,4-dien-3-one (20). Prep- aration from 11 (200 mg, 0.46 mmol) and NaSC2H, according to method A gave 58 mg (26%) of thiol ester 20 and 92 mg (46%) of returned 11. The analytical sample was recrystallized twice from acetone-hexane: mp >300 "C; UV 238 nm (t 19595); IR (KBr) 1745 (ester), 1665 (br, 20-C=0,3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 0.90 (d, 3 H, J = 7 Hz, 16-CH3), 0.98 (s, 3 H, 18-CH3), 1.16 (t, 3H, J = 7 Hz, CH,CH,), 1.46 (8 , 3 H, Ig-CH,), 1.98 (s, 3 H, COCH,), 2.84 (9, 2 H, J = 7 Hz, SCH,); MS, m / e 421 (M+ - SC,H5). Anal. Calcd for CZ5H3,O5F2S: C, 62.23; H, 6.68. Found: C, 62.50; H, 6.63. 6a,9a-Difluoro-l la-hydroxy-l7/3-[ (methy1thio)-

carbonyl]-17a-(propionyloxy)-l6a-methylandrosta-l,4-dien- %one (21). The reaction of 12 (1.5 g, 3.32 mmol) with NaSCH, according to method A gave 376 mg (25%) of returned 12 and 525 mg (33%) of thiol ester 21 in two crops. Analytical sample from acetone-hexane: mp 285-287 "C dec; UV 238 nm (t 19565); IR (KBr) 1740 (ester), 1675 (20-C=0), 1665 (3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 0.89 (d, 3 H, J = 7 Hz, 16-CH3), 0.93 (s, 3 H, 18-CH3), 1.00 (t, 3 H, J = 7 Hz, CH,CH,), 1.46 (8 , 3 H,

MS, m / e 482 (M+). Anal. Calcd for C25H3205F2S: C, 62.22; H, 6.68; S, 6.64. Found: C, 62.45; H, 6.82; S, 6.59. Method B proved superior for the preparation of 21. Thus, the reaction of 12 (200 mg, 0.44 mmol) with diphenyl chlorophosphate (168 pL, 0.81 mmol) and TEA a t 55 "C followed by treatment with NaSCH3/DMF (0.83 mmol) gave 189 mg (89%) of 21, mp 286-288 "C dec. 17a-Acetoxy-6a,9a-difluoro- 1 la-hydroxy- 16a-methyl-3-

oxoandrosta-1,4-diene-l7j3-carboxylic Diethyl Phosphoric Anhydride (22). Diethyl chlorophosphate (120 pL, 0.83 mmol) was added to a solution of 11 (219 mg, 0.50 m o l ) in THF (6 mL)

1.48 (s, 3 H, 19-CH3), 2.01 (9, 3 H, COCH,), 2.26 (s, 3 H, SCH,);

19-CH3), 2.24 (9, 3 H, SCH,), 2.29 (4, 2 H, J = 7 Hz, COCH,);

J. Org. Chem., Vol. 51, No. 12, 1986 2325

containing TEA (125 pL, 0.9 mmol) and the mixture was warmed at 50 "C. After 1 h, TLC (0.5% HAC-35% acetone-hexane) showed that all but a trace of 11 had been converted to the intermediate ( R f 0.32) seen in preparations of 19 by method A. The mixture was cooled, diluted with EtOAc, washed with dilute Na2C03 and water, dried (Na2S04), and evaporated to dryness. The residue was partially purified by dissolving in acetone (2 mL) containing a drop of TEA, adding hexane (20 mL), and decanting the supernatant. Vacuum drying of the precipitate gave 242 mg (84%) of 22 as an amorphous solid judged >95% pure by TLC. An analytical sample was purified by centrifugal TLC (0.1% TEA-20% acetone-CHzClz) and precipitation as before, affording a semi-crystalline solid, mp 100-110 "C dec; W 238 nm (t 16 100); IR (CHCI,) 1780 (20-C=0), 1745 (ester), 1670 (3-C-0) cm-'; 'H NMR (300 MHz, CDC1,) 6 0.96 (d, 3 H, J = 7 Hz, 16-CH3), 1.17 (s, 3 H, 18-CH3), 1.36 (dt, 3 H, J1 = 1 Hz, 5, = 7 Hz, CH,CH,),

4.35 (dq, 2 H, J1 = 7 Hz, 5, = 7 Hz, OCH,); MS, m / e 575 (MH'). Anal. Calcd for C27H370$'2P: C, 56.44; H, 6.49. Found: C, 56.53; H, 6.57. Samples of 22 stored under argon at -20 "C remained 95-98% pure for 1-2 months, after which degradation (largely to 11) was observed.

l7a-Acetoxy-9a-fluoro-l1@-hydroxy-17~-[ (methy1thio)- carbonyl]-16~-methylandrosta-1,4-dien-3-one (24) and 1 1~,17a-Dihydroxy-Sa-fluoro- 16/3-met hyl-3-oxoandrosta- 1,4-diene-l7&carboxylic Acid 17,20-(l'-Methyl-l'-(methylthio))methylene Ketal (27). The activation of 13 (168 mg, 0.40 mmol) by method B and reaction with NaSCH, (0.60 mmol) afforded (after recrystallization from acetone-hexane) 134 mg (74%) of 24, mp 263-265 "C dec; UV 239 nm (t 19600); IR (KBr) 1745 (ester), 1700 (20-C=0), 1660 (3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 0.86 (s, 3 H, 18-CH3), 1.28 (d, 3 H, J = 7 Hz, 16-CH3), 1.49 (s, 3 H, 19-CH3), 1.98 (s, 3 H, COCH,), 2.14 (s, 3 H, SCH,); MS, m/e 403 (M+ - SCH,). Calcd for C24H3105FS: C, 63.98; H, 6.94; S, 7.11. Found: C, 63.73; H, 7.07; S, 7.41. The reaction of 13 (210 mg, 0.50 mmol) with diethyl chlorophosphate (0.70 mmol) and NaSCH, (0.70 mmol) according to method A gave 180 mg (80%) of a mixture, shown by TLC (0.25% TEA-55% ether-cyclohexane, developed three times) to be approximately 9 1 of ortho ester 27 (Rf 0.38) with 24 (Rf 0.44). The tendency of 27 to isomerize to 24 on silica gel during chromatography was partially overcome by inclusion of TEA in the eluent. Thus, a sample of 27 was purified for analysis by centrifugal TLC (0.4% TEA-65% ether-cyclohexane, developed 3 times) and recrystallization from ether-pentane: mp 202-203 "C dec; UV (e 16000); IR (KBr) 1795 (20-C=0), 1665 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 1.15 (d, 3 H, J = 7 Hz, 16-CH3), 1.17 (s, 3 H, 18-CH3), 1.51 (s, 3 H, 19-CH3), 1.83 (s, 3 H, 1'-CH,), 2.19 (s,3 H, SCH,); MS, m / e 451 (MH'). Anal. Calcd for C24H310&3: C, 63.98; H, 6.94. Found: C, 63.86; H, 7.03.

17a-Acetoxy-9a-fluoro- 1784 (n -hexylthio)carbonyl]-ll~- hydroxy-16/3-methylandrosta-1,4-dien-3-one (25) and 11~,17a-Dihydroxy-9a-fluoro-16~-methy1-3-oxoandrosta- l,4-diene- 17j3-carboxylic Acid 17,20-( 1'-Methyl-1'-(n -hexylthio))methylene Ketal (28). The reaction of 13 (220 mg, 0.5 mmol) with diethyl chlorophosphate and NaSC6H13 according to method A and separation of the product mixture by preparative TLC (4% acetone-CHzCl2, developed twice) afforded 35 mg (13%) of thiol ester 25 (R, 0.38), mp 168-172 "C, and 111 mg (43%) of ortho thio ester 28 (Rf 0.33). A sample of 25 was recrystallized from EtOAc-hexane for analysis: mp 176-178 "C; UV 240 nm (c 19025); IR (KBr) 1730 (ester), 1700 (20-C=0), 1660 (3-C=0) cm-'; 'H NMR (90 MHz, Me2SO) 6 0.87 (br m, 6 H, 18-CH,, CHzCH3), 1.28 (br m, 11 H, 16-CH3, hexyl CH2's), 1.49 (s, 3 H, 19-CH3), 2.00 (s,3 H, COCH,); MS, m / e 403 (M' - SCsH13). Anal. Calcd for C29H4105FS: C, 66.89; H, 7.94. Found: C, 66.82; H, 8.07. The data for 28 (from acetone-hexane) were as follows: mp 129-131 "C; UV 237 nm (t 16310); IR (KBr) 1795 (20-C=0), 1675 (3-C=0) cm-', 'H NMR (90 MHz, Me,SO) 6 0.85 (br t, 3 H, J = 5 Hz, CH,CH,), 1.14 (d, 3 H, J = 7 Hz, 16-CH3), 1.16 (s, 3 H, l&CH3), 1.50 (s, 3 H, 19-CH3), 1.84 (s, 3 H, l'-CH3); MS, m/e 403 (M+ - SC8H13). Anal. Calcd for C29H4105FS: C, 66.89; H, 7.93. Found: C, 67.01; H, 8.00. Sa-Fluoro-1 lj3-hydroxy-17,9-[ (methylthio)carbonyl]-17a-

(propionyloxy)-16~-methylandrosta-l,4-dien-3-one (26) and

1.39 (dt, 3 H, 51 = 1 Hz, Jz = 7 Hz, CHZCHJ, 1.54 (s, 3 H, Ig-CH,), 2.10 ( ~ , 3 H, COCH,), 4.21 (dq, 2 H, 51 = 7 Hz, 52 = 7 Hz, OCHZ),

Anal.

2326 J. Org. Chem., Vol. 51, No. 12, 1986

the Isomeric 1790-( 1'-Ethyl- 1'-methy1thio)methylene Ketal (29). The activation of acid 14 (113 mg, 0.26 mmol) with diethyl chlorophosphate (1 mmol) and treatment with NaSCH, (2.5 mmol) according to method A gave 133 mg of a neutral residue which was purified by preparative TLC (2% CH30HCH2C12, developed twice). The product was 77 mg (64%) of a chromatographically homogeneous solid, estimated to be a 3:l mixture of ketal 29 with thiol ester 26 by comparing the integrals of the methylthio group 'H NMR resonances (2.12 ppm for 29, 2.17 ppm for 26) and intensities of the major infrared absorptions (1790 for 17P-carbonyl of 29,1740 cm-' for acetate of 26). A sample of the mixture (64 mg, 0.14 mmol) in THF (8 mL) was treated with additional NaSCH,/DMF (1.65 mmol), and the mixture was heated a t 45 "C for 6 h. Acid-base workup and recrystallization from acetone-hexane then afforded 30 mg (30% overall yield) of pure 26: mp 223-224 "C; UV 239 nm (t 18890); IR ( O r ) 1740 (ester), 1700 (20-C=0), 1665 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.88 (9, 3 H, 18-CH3), 1.03 (t, 3 H, J = 7 Hz, CHzCH3), 1.31 (d, 3 H, J = 7 Hz, 16-CH3), 1.50 (s,3 H, 19-CH3), 2.17 (s, 3 H, SCH3), 2.34 (q, 2 H, J = 7 Hz, COCH,); MS, m/e 464 (M+). Anal. Calcd for C26H3305FS: C, 64.63; H, 7.16; S, 6.90. Found: C, 64.38; H, 7.34; S, 6.87. Thiol ester 26 was also prepared by reaction of 14 (5.7 g, 13.26 mmol) with diphenyl chlorophosphate (19.3 mmol) and NaSCH3 (20.8 mmol) according to method B. Analysis of the reaction mixture after 1 h a t 20 "C, indicated the product to be 26 containing 510% of 29. Extra NaSCH, (5 mmol) was added and the mixture was heated at 60 "C for 2 h, after which recovery and recrystallization of the product afforded 3.9 g (64%) of pure 26, mp 223-224 "C. A third sample of 26 (34%) was prepared via activation of 14 with thionyl chloride, using the procedure described for 19. 6a,9a-Difluoro-11@,17a-dihydroxy- 16a-methyl-3-oxo-

androsta-1,4-diene-l7@-carboxylic Acid 17,204 1'-Methyl- 1'- methoxy)methylene Ketal (30). A solution of freshly prepared 22 (104 mg, 0.18 mmol) in MeOH (10 mL) was heated at reflux for 1 h, giving a 1:l mixture of ketal isomers by TLC (Ris 0.30, 0.37,40% EtOAc-hexane). The solution was diluted with EtOAc, washed with dilute NaHC0, and water, dried (Na2S04), and evaporated to dryness. Recrystallization from acetone-hexane afforded 58 mg (71%) of 30, mp 200-202 "C dec; UV 238 nm ( e 16300); IR (KBr) 1795 (20-C=0), 1670 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.99, 1.04 (2 d, 3 H, J's = 7 Hz, 16-CH3's);

(2 s, 3 H, 1'-CH3's), 3.39, 3.42 (2 s, 3 H, OCH,'s); MS, m/e 452 (M'). Anal. Calcd for Cz4H,06Fz: C, 63.71; H, 6.68. Found: C, 63.89: H, 6.64. A second sample of 30 was prepared from 11 (175 mg, 0.40 mmol) by reaction with diethyl chlorophosphate (0.56 mmol) and TEA (0.72 mmol) in THF according to method A followed by filtration of the mixture into a flask containing MeOH ( 5 mL). The THF was evaporated and the solution was stirred at 20 "C for 16 h and at 50 "C for 1 h, after which isolation as above afforded 154 mg (85%) of 30, mp 198-200 "C.

11~,17a-Dihydroxy-9a-fluoro-16~-methy1-3-oxoandrosta- l,a-diene- 17@-carboxylic Acid 17,20-( 1'-Methyl- 1'-met h- 0xy)methylene Ketal (31). Following activation of acid 13 (168 mg, 0.40 mmol) with diethyl chlorophosphate (81 pL, 0.56 mmol) and TEA (100 pL, 0.72 mmol) and exchange of the THF solvent for MeOH as described above, the formation of 31 was complete in 16 h at 20 "C. The ketal isomers (Rf's 0.37) were not separable by TLC (40% EtOAc-hexane). Recovery and recrystallization of the product as described for 30 afforded 132 mg (76%) of 31: mp 223-225 "C dec; UV 238 nm ( e 15400); IR (KBr) 1790, (20- H), 1665 (3-C=0) cm-'; 'H Nh4R (300 MHz, CDCl,) 1.20, 1.22

1.30, 1.32 (2 S, 3 H, 18-CH3'~), 1.54 (s, 3 H, 19-CH3), 1.71, 1.72

(2 d, 3 H, J'S = 7.5 Hz, 16-CH,'s), 1.27, 1.31 (2 8, 3 H, 18-CH,'s), 1.57 (9, 3 H, 19-CH3), 1.65, 1.66 (2 S, 3 H, l'-CH,'s), 3.31, 3.32 (2 s, 3 H, OCH,'s); MS, m/e 435 (MH'). Anal. Calcd for C24H31O8F: C, 66.34; H, 7.19. Found: C, 66.42; H, 7.34.

Reactions of Acids 6-9,11, and 13 wi th FMPT: Method D. Preparations were carried out in CH2Cl2 containing TEA as described in method C, except that thiolate was not added. Reactions were worked up after 30 min to 2 h at -15 "C.

GqSa-Difluoro- 11~,17a-dihydroxy-16a-methyl-3-oxoandrosta- 1,4-diene-17j3-carboxylic Acid 17,20-( 1'-Methyl-1'- fluoro)-(S)-methylene Ketal (32) and Isomeric @)-Methyl Ketal (33). The treatment of 11 (219 mg, 0.50 mmol) with FMPT at -15 "C: (method D) resulted in conversion to two less polar

Kertesz and Marx

products (Rfs 0.40,0.57,1% HOAc-35% EtOAc-hexane). The reaction mixture was diluted with CH2C1, (50 mL), washed with dilute Na2C03 and water, dried (NaZSO4), and evaporated to dryness, after which centrifugal TLC (45% EtOAc-hexane) and recrystallizations from acetone-hexane afforded 74 mg of ketal 32 (34%) and 66 mg of less polar isomer 33 (30%). The 13-Hz coupling of the 1'-methyl group in the 'H NMR spectrum of each compound was consistent with the presence of a vicinal fluorine, and the additional 2-Hz coupling of the 16-CH3 by a proximal fluorine was the basis for assignment of "S" stereochemistry to ketal 32.28 The data for 32 were as follows: mp 213-214 "C dec; UV 237 nm ( e 15770); IR (KBr) 1810 (20-C=0), 1665 (3-C=0) cm-'; 'H NMR (100 MHz, CDCl,) 6 1.03 (dd, 3 H, J1 = 7 Hz, J2 = 2 Hz, 16-CH3), 1.30 (s, 3 H, WCH,), 1.52 (8, 3 H, 19-CH,), 1.82 (d, 3 H, J = 13, 1'-CH,); MS, m/e 440 (M+). Anal. Calcd for C23H2705F3: C, 62.72; H, 6.18. Found: C, 62.67; H, 6.19. The data for 33 were as follows: mp 210-211 "C dec; UV 236 nm ( t

15570); IR (KBr) 1805 (20-C=0), 1670 (3-C=0) cm-'; 'H NMR (100 MHz, CDCl,) 6 0.95 (d, 3 H, J = 7 Hz, 16-CH3), 1.31 (s, 3 H, 18-CH3), 1.52 (9, 3 H, 19-CH3), 1.90 (d, 3 H, J = 13, l'-CH3); MS, m/e 440 (M+). Anal. Calcd for CZ3Hz7O5F3: C, 62.72; H, 6.18. Found C, 62.53; H, 5.94. Attempts to convert fluoroketals 32 and 33 in a freshly prepared reaction mixture into thiol ester 19 failed. No reaction was observed following the addition of an excess of methanethiol in methylene chloride containing TEA, followed by addition of a 4-fold excess of 0.83 N NaSCH,/DMF and warming at a reflux, followed by exchange of the solvent for toluene and heating at reflux for 16 h.

11@,17a-Dihydroxy-9a-fluoro-l6@-methyl-3-oxoandrosta- 1,4-diene-l7&carboxylic Acid 17,204 1'-Methyl-1'-fluoro)- methylene Ketal (34) and Isomeric Ketal (35). Treatment of 13 (210 mg, 0.50 mmol) with FMPT (method D) gave a 5:1 mixture of two neutral products (R,'s 0.42, 0.49, 40% EtOAc- hexane, developed twice). Workup and two recrystallizations of the crude product from acetone-hexane afforded 85 mg (40%) of less polar fluoroketal34, mp 185-186 "C dec; UV 238 nm ( e 15700); IR (CHCl,) 1810 (20-C=0), 1665 (3-c=O) cm-'; 'H NMR (300 MHz, CDC1,) 6 1.19 (d, 3 H, J = 7.5 Hz, 16-CH3), 1.26 (s, 3 H, 18-CH3), 1.56 (s, 3 H, 19-CH3), 1.80 (d, 3 H, J = 13 Hz, l'-CH3); MS, m/e 402 (M+ - HF). Anal. Calcd for Cz3H2,05Fz: C, 65.39; H, 6.68. Found: C, 65.53; H, 6.69. The mother liquor materials were recovered and purified by centrifugal TLC (0.2% TEA-40% EtOAc-hexane), and the separated products were recrystallized from acetone-hexane, which gave an additional 35 mg (17%) of pure 34 and 21 mg (10%) of more polar fluoro ketal 35, mp 193-194 "C dec; UV 238 nm ( t 15300); IR (CHCl,) 1810 (ZO-C===U), 1670 (3-C=0) cm-'; 'H NMR (300 MHz, CDC1,) 6 1.21 (d, 3 H, J = 7.5 Hz, 16-CH3), 1.30 (9, 3 H, 18-CH3), 1.56 (s, 3 H, 19-CH3), 1.78 (d, 3 H, J = 13 Hz, l'-CH3); MS, m/e 402 (M+ - HF). Anal. Calcd for CZ3Hz8O5F2: C, 65.39; H, 6.68. Found: C, 65.47; H, 6.40. Ketals 34 and 35 appeared more labile on silica gel than the 16a-methyl analogues, which necessitated the inclusion of TEA in the chromatographic eluent. 9a,l1~-Dichloro-6a-fluoro-16-methyl-3-oxoandrosta-

1,4,16-triene-17-carboxylic Acid (36). The preparation of thiol ester 47 from acid 9 (180 mg, 0.42 mmol) was attempted by method C. However, the major less polar product of reaction with FMPT (R, 0.38, 1% HOAc-25% EtOAc-hexane, developed twice) remained unchanged after addition NaSCH,/DMF (1.41 mmol) and warming to 20 "C, and workup then afforded 58 mg (34%) of Alfi-acid 36 and a variety of minor neutral materials which were not further investigated. Analytical sample from acetonehexane: mp 252-254 "C dec; UV 234 nm ( e 21040); IR (KBr) 1660 (br, COOH, 3-C=0) cm-'; 'H NMR (90 MHz, MezSO) 6 1.27 (s, 3 H, 18-CH3), 1.74 (s, 3 H, Ig-CH,), 2.04 (s, 3 H, 16-CH,), 4.98 (br m, 1 H, 11-H); MS, m/e 412,414 (M'). Anal. Calcd for CzlHZ3O3Cl2F: C, 61.03; H, 5.61. Found: C, 61.12; H, 5.54.

Ga,Sa-Difluoro- 1 l@,l7a-dihydroxy- 178-[ (met hy1thio)- carbonyl]-l6a-methylandrosta-l,4-dien-3-one (37). The reaction of 6 (198 mg, 0.05 mmol) with FMPT and NaSCH3 (method C) gave a mixture of three neutral products and 14 mg (7 %) of returned 6. Separation of the neutral components by centrifugal TLC (6% acetone-CH2Cl2) afforded 63 mg (30%) of 37 (R, 0.25), along with 31 mg (18%) of triene 38 (Rf0.40) and 22 mg (12%) of lactone 39 (Rf 0.15) (vide infra). Analytical sample of 37 from acetone-hexane: mp 275-277 "C dec; UV 239 nm (t 19550); IR

Thiol Esters from Steroid 176-Carboxylic Acids

(KBr) 1680 (20-C=0), 1665 (3-C=0) cm-'; 'H NMR (300 MHz, Me2SO) 6 0.83 (d, 3 H, J = 7 Hz, 16-CH3), 0.92 (s, 3 H, 18-CH3), 1.49 (s, 3 H, 19-CH3), 2.17 (s, 3 H, SCHJ; MS, m / e 426 (M'). Anal. Calcd for Cz2HaO4FzS: C, 61.95; H, 6.62. Found: C, 61.98; H, 6.74. 6~,9a-Difluoro-16a,l7-dimethyl-l lg-hydroxy-18-nor-

androsta-1,4,13(17)-trien-3-one (38) and Ga,Sa-Difluoro- 11~,13a-dihydroxy-16a,l7~-dimethyl-3-oxo-l8-norandrosta- 1,4-diene-17a-carboxylic Acid 13-Lactone (39). The reaction of 6 (158 mg, 0.40 mmol) with FMPT according to method D followed by workup, preparative TLC, and recrystallizations from acetone-hexane gave 22 mg (16%) of 38 and 17 mg (11%) of 34 along with 11 mg (7%) of recovered 6 as the only isolable products. The data for 38 were as follows: mp 215-217 "C; UV 234 nm ( e 15700); IR (KBr) 1670 (3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 0.92 (d, 3 H, J = 7 Hz, 16-CH3), 1.39 (s,3 H, 17-CH3), 1.53 (s, 3 H, 19-CH3); MS, m / e 334 (M'). Anal. Calcd for CzoH2402F2: C, 71.83; H, 7.23. Found: C, 71.70; H, 7.05. The data for 39 were as follows: mp 154-156 "C dec; UV 236 nm ( e 15800); IR (CHCl,) 1815 (20-C=0), 1670 (3-C=O) cm-'; 'H NMR (300 MHz, CDCl,, trace CD,OD) 6 1.14 (d, 3 H, J = 7.5 Hz, 16-CH3), 1.36 (s, 3 H, 17-CH3), 1.43 (s, 3 H, 19-CH3); 13C NMR (CDCl,, trace CD,OD) 6 13.15, 13.19 (16-CH3, 17-CH3), 23.73 (Ig-CH,), 40.28 (C-16), 45.35 (C-14), 66.24 (C-17),66.76 (d, JCF = 0.40 Hz, C-ll), 86.45 (d, JCF = 2.45 Hz, C-6), 90.57 (C-13), 97.54 (d, JCF = 2.39 Hz, C-9), 173.61 (17-C=0), 185.73 (C-3); MS, m / e 378 (M'); HRMS, m / e calcd for CZ1Hz4O4F2 378.1643, found 378.1640. A satisfactory elemental analysis could not be obtained. A sample of 39 which had been dried in vacuo at 100 "C for 16 h was completely transformed into the triene 38 (presumably by loss of Cod, as confirmed by TLC, loss of the 1815 cm-' absorption in the infrared, and the 'H NMR spectrum. Samples of 39 were subsequently dried at ambient temperature and proved to be reasonably stable when stored at -20 "C for up to 2 months, after which significant degradation to 38 was observed. 16~,17-Dimethyl-9a-fluoro-l1~- hydroxy-18-norandrosta-

1,4,13( 17)-trien-3-one (40) and llj3,13a-Dihydroxy-l68,178- dimethyl-9a-fluoro-3-oxo- 18-norandrosta- 1,a-diene- 17a- carboxylic Acid 13-Lactone (41). Reaction of 7 (151 mg, 0.40 mmol) with FMPT (method D) gave 107 mg of a neutral product which proved to be a mixture of triene 40 (Rf 0.44) and the lactone 41 (Rf 0.24) by TLC (7% acetone-CHzC12). %crystallization from acetone-hexane gave 30 mg (24%) of pure 40, and centrifugal TLC of the mother liquor materials (same system) followed by recrystallizations from acetone-hexane gave a further 31 mg (24%) of 40 and 12 mg (8%) of 41. Products and yields were not altered by addition of NaSCH, to the reaction mixture in a separate experiment. Data for 40 were as follows: mp 256-258 "C; UV 236 nm ( e 20070); IR (KBr) 1660 (3-C=0) cm-'; 'H NMR (100 MHz, Me2SO) 6 1.00 (d, 3 H, J = 7 Hz, 16-CH3), 1.39 (s, 3 H, 17-CH3), 1.51 (s, 3 H, 19-CH3); MS, m / e 316 (M'). Anal. Calcd for C20H2502F: C, 75.92; H, 7.96. Found: C, 75.77; H, 7.92. Lactone 41 was reasonably stable in solution, but dry samples were more labile than the 16a-methyl analogue 39 and rapidaly degraded to triene 40 on storage. The 13C NMR resonances for the 16- and 17-methyls of 41 were shifted related to those of lactone 39, and were consistent with the assigned cis-dimethyl system. Data for 41 were as follows: mp 258-260 "C dec (from CH2C12-hexane); UV 237 nm (C 15200); IR (CHCl,) 1815 (20- C=O) 1670 cm-' (3-C=0); 'H NMR (300 MHz, CDCl,) 6 1.15 (d, 3 H, J = 7.5 Hz, 16-CH3), 1.34 (s, 3 H, 17-CH3), 1.45 (s, 3 H, 19-CH3); 13C NMR (CDCl,) 6 11.89, 17.97 (16-CH3, 17-CH3), 23.63 (N-CH,), 28.95 (C-6), 37.89 (C-16), 45.79 (C-14), 65.82 (C-17), 67.97 (d, JCF = 0.42 Hz, C-ll) , 91.55 (C-13), 98.19 (d, JCF = 2.38 Hz, C-9), 175.59 (17-C=0), 185.99 (C-3); MS, m / e 340 (M' - HF); HRMS, m / e calcd for C21H2404 (M+ - HF) 340.1675, found 340.1699, calcd for CzoHzs02F (M' - COz) 316.1839, found 316.1826. 6a-Fluoro-l6-methyl-3-oxoandrosta-1,4,9( 1 l),ls-tetraene-

17-carboxylic Acid (42) and 6a-Fluoro-l7-[(methylthio)- carbonyl]-l6-methylandrosta-l,4,9( 11),16-tetraen-3-one (43). Reaction of 8 (170 mg, 0.47 mmol) with FMPT and NaSCH, (0.83 mmol) (method C) gave 77 mg of acidic material and 104 mg of neutral residue as an oil. Two crystallizations of the acid fraction from EtOAc-hexane gave 50 mg (30%) of 42, mp 27Ck271 "C dec; W 235 nm ( e 21000); IR (KBr) 1715 (COOH), 1670 (3-C=0) cm-';

J . Org. Chem., Vol. 51, No. 12, 1986 2327

'H NMR (100 MHz, Me2SO) 6 0.84 (s, 3 H, 18-CH,), 1.36 (s, 3 H, 19-CH3), 2.01 (s, 3 H, 16-CH3), 5.57 (br d, 1 H, J = 4 Hz, 11-H); 13C NMR (CDCl,) 6 15.47 (18-CH3), 17.48 (16-CH,), 26.88 (19- CHJ, 38.27 (C-15), 40.12 (C-12), 40.35 (C-7), 45.85 (C-13), 50.93 (C-14), 87.31 (d, JcF = 2.44 Hz, C-6), 119.66 (C-4), 122.74 (C-ll), 127.75 (C-2), 135.67 (C-17), 141.23 (C-g), 158.79 (C-l6), 161.89 (C-5), 170.19 (17-C=0), 185.39 (C-3); MS, m / e 342 (M'). Anal. Calcd for C21HDO3F C, 73.66; H, 6.77. Found: C, 73.54; H, 6.67. A difficult purification of the neutral residue by preparative TLC (15% acetonehexane, developed five times) and recrystallzation of the very soluble major component from pentane gave 26 mg (15%) of thiol ester 43, mp 259-262 "C; UV 240 nm (C 18005); IR (KBr) 1700-1665 cm-' (br, 20-C=0,3-C=0); 'H NMR (100 MHz, Me2SO) 6 0.94 (s, 3 H, 18-CH3), 1.37 (s, 3 H, 19-CH3), 2.04 (s,3 H, 16-CH3), 2.25 (s,3 H, SCH3), 5.57 (br d, 1 H, J = 6, 11-H); MS, m / e 372 (M'). Anal. Calcd for CZ2Hz5O2FS: C, 70.94; H, 6.76. Found: C, 70.98; H, 6.50. 6a-Fluoro-l6-methyl-3-oxoandrosta- 1,4,9( 11),16-tetraene-

17-carbonyl Fluoride (44). The reaction of 8 (190 mg, 0.50 mmol) with FMPT alone (method D) gave 48 mg (28%) of acid 42 and 35 mg of a neutral fraction containing a single major product (Rr 0.58,4% acetone-CH2C12). Centrifugal TLC (same system) of the labile neutral product and recrystallization from acetone-hexane gave 11 mg (8%) of 44, mp 243-244 "C dec; UV 237 nm (C 23400); IR (CHCl,) 1790 (20-C=0), 1670 (3-C==0) cm-'; 'H NMR (300 MHz, CDC1,) 6 0.94 (s, 3 H, WCH,), 1.42 (s, 3 H, 19-CH3), 2.19 (s, 3 H, 16-CH3), 5.64 (d, 1 H, J = 6.6 Hz, 11-H); MS, m / e 344 (M+); HRMS, m / e calcd for CzlHzzOzF2 344.1588, found 344.1577. Anal. Calcd for C21H2202F2: C, 73.24; H, 6.44. Found: C, 73.12; H, 6.50. 6a,9a-Difluoro-l1~,17a-dihydroxy-l7~-[ (ethy1thio)-

carbonyl]-16a-methylandrosta-1,4-dien-3-one (45). To a solution of 6 (105 mg, 0.27 mmol) in DMF (7 mL) at -10 "C was added CDI (80 mg, 0.49 mmol) in DMF (3 mL) and the mixture was stored at -5 "C for 16 h, after which conversion to a slightly less polar material was complete (TLC, 0.25% HOAc-40% acetone-hexane). Ethyl mercaptan (0.2 mL, 2.7 mmol) was added, and the mixture was stirred at 20 "C for 16 h. The solution was evaporated to dryness, and the residue was purified by preparative TLC (10% acetone-benzene) to give 47 mg (40%) of 45. Ana- lytical sample, recrystallized 4 times from acetone-hexane: mp 253-256 "C dec; UV 238 nm ( e 19730); IR (KBr) 1665 cm-' (br, 20-C=0,3-C=0); 'H NMR (100 MHz, Me2SO) 6 0.81 (d, 3 H, J = 7 Hz, 16-CH3), 0.90 (s, 3 H, 18-CH3), 1.14 (t, 3 H, J = 7 Hz,

MS, m / e 440 (M+). Anal. Calcd for C23H3004F2S: C, 62.71; H, 6.86. Found: C, 62.49; H, 7.08.

11~,17a-Dihydroxy-9a-fluoro-17@-[ (me thy1 th io ) - carbonyl]-16~-methylandrosta-1,4-dien-3-one (46). An ice- cooled solution of 7 (100 mg, 0.26 mmol) and CDI (80 mg, 0.49 mmol) in DMF (8 mL) was stirred for 2 h, after which a gentle stream of CH3SH gas was introduced below the surface of the mixture for 1 h. The mixture was purged with nitrogen and evaporated to dryness, and the product was purified by centrifugal TLC (3% MeOH-CH2C12) and recrystallization from acetone- hexane to give 66 mg (61%) of pure 46, mp 242-244 "C dec; UV 239 nm ( e 19400); IR (KBr) 1690 (20-C=0), 1670 (3-C=0) cm-l; 'H NMR (100 MHz, Me2SO) 6 0.95 (s, 3 H, 18-CH3), 1.02 (d, 3 H, J = 7 Hz, 16-CH3), 1.47 (s, 3 H, 19-CH3), 2.04 (s, 3 H, SCH,); MS, m / e 388 (M+ - HF). Anal. Calcd for CzzHz9O4FS: C, 64.68; H, 7.16. Found: C, 64.69; H, 7.20.

9a, 1 1~-Dichloro-6a-fluoro- 17a-hydroxy- 178- [(methylthio)carbonyl]-16~-methylandrosta-1,4-dien-3-one (47). Reaction of 9 (200 mg, 0.46 mmol) with CDI and CH3SH as described for the preparation of 46 and recrystallization of the product from acetone-hexane gave 180 mg (84%) of 47, mp 232-233 "C dec; UV 237 nm ( e 19850); IR (KBr) 1675 cm-' (20- C=O, 3-C=0); 'H NMR (90 MHz, Me2SO) 6 1.03 (s, 3 H, 18- CH,), 1.07 (d, 3 H, J = 7 Hz, 16-CH3), 1.70 (s, 3 H, 19-CH3), 2.10 (s,3 H, SCH,), 4.99 (br d, 1 H, J = 4 Hz, 11-H); MS, m / e 461-465 (MH+). Anal. Calcd for C22H270,C12FS: C, 57.27; H, 5.90; C1, 15.37. Found: C, 57.34; H, 5.93; C1, 15.33. 17a-Acetoxy-9a,l1@-dichloro-6a-fluoro-l7~-[ (methyl-

thio)carbonyl]-16~-methylandrosta-1,4-dien-3-one (48). A solution of 47 (110 mg, 0.24 mmol) and DMAF' (15 mg, 0.12 mmol) in TEA (2 mL) and acetic anhydride (2 mL) was heated at 80 "C

CHZCH,), 1.46 (9, 3 H, Ig-CH,), 2.72 (9, 2 H, J = 7 Hz, SCH2);

2328 J . Org. Chem. 1986,51, 2328-2332

for 1.5 h and then stirred a t 20 "C for 16 h. The mixture was diluted with EtOAc, washed with water, and evaporated to dryness. The product was purified by preparative TLC (20% acetone-hexane, eluted 4 times) and crystallization from acetone- hexane, giving 57 mg (47%) of 48, mp 250-251 'C dec; UV 235 nm (c 19 345); IR (KBr) 1750 (ester), 1700 (20-C=O), 1675 (3- C=O) cm-'; 'H NMR (90 MHz, Me2SO) 6 0.95 ( s ,3 H, 18-CH3), 1.34 (d, 3 H, J = 7 Hz, 16-CH3), 1.71 (s, 3 H, 19-CH3), 2.05 (s, 3 H,COCH3), 2.20 (8, 3 H, SCH3), 5.09 (br d, 1 H , J = 4 Hz, 11-H); MS, m / e 503-507 (MH'). Anal. Calcd for C24Hzs04C12FS: C, 57.26; H, 5.81. Found: C, 57.19; H, 5.85.

9a, 11~-Dichloro-6a-fluoro- 1784 (methylthio)carbonyl]- 17a-(propionyloxy)-16~-methylandrosta-l,4-dien-3-one (49). Freshly distilled propionic anhydride (0.3 mL, 2.34 mmol) was added to a solution of 47 (140 mg, 0.30 mmol) and DMAP (50 mg, 0.41 mmol) in TEA (3 mL), and the mixture was heated at 70 "C for 16 h. Purification of the product by centrifugal TLC (0.5% acetone-CH2C12) followed by crystallizations from MeOH and EtOAc-hexane gave 55 mg (35%) of 49, mp 244-246 "C dec; UV 237 nm ( e 18995); IR (Kl3r) 1745 (ester), 1705 ( 2 0 - M ) , 1670 (3-C=0) cm-I; 'H NMR (300 MHz, Me2SO) 6 0.95 (s, 3 H, 18-

16-CH3), 1.71 (8 , 3 H, 19-CH3), 2.19 (s, 3 H, SCH3), 2.33 (4, 2 H, J = 7 Hz, COCHJ; MS, m/e 468-472 (M+ - HSCH3). Anal. Calcd for C25H3104C12FS: C, 58.03; H, 6.04; C1, 13.70. Found: C, 57.98; H, 6.04; C1, 13.68.

Acknowledgment. We thank Dr. F. S. Alvarez,5l Dr.

CH3), 1.05 (t, 3 H, J = 7 Hz, CH&H3), 1.33 (d, 3 H, J = 7 Hz,

J. Patterson, and Paul Wagner for their contributions t o this work. We are indebted to Janis Nelson and Dr. M. Maddox for the measurement and interpretation of NMR spectra, and to Dr. I. J. Massey and Dr. L. Partridge for mass spectra. We also thank Dr. A. F. Kluge and Dr. K. A. M. Walker for many helpful discussions and suggestions.

Registry No. 1,67-73-2; 2, 2135-17-3; 3, 378-44-9 4, 101916- 22-7; 4 diacetate, 50630-16-5; 5, 101916-23-8; 5 diacetate, 101916-49-8; 6, 28416-82-2; 7, 37926-75-3; 8, 101916-24-9; 9, 101916-25-0; 10,65751-34-0; 11,101916-26-1; 12,65429-42-7; 13, 37927-21-2; 14, 37927-23-4; 15, 101916-27-2; 16, 37927-22-3; 17, 74131-78-5; 18, 101916-28-3; 19, 101916-29-4; 20, 74131-73-0; 21, 73205-13-7; 22, 101916-30-7; 23,101916-31-8; 24,79578-14-6; 25, 74156-42-6; 26,79578-08-8; 27, 101932-58-5; 28,101916-32-9; 29, 101916-33-0; 30 (isomer l ) , 101916-34-1; 30 (isomer 2), 101916-35-2; 31 (isomer l), 101916-36-3; 31 (isomer 2), 101916-37-4; 32, 101932-54-1; 33, 101916-38-5; 34, 101916-39-6; 35, 101916-40-9; 36,101916-41-0; 37, 74131-77-4; 38, 101916-42-1; 39,101916-43-2; 40,101916-44-3; 41,101932-61-0; 42,101932-59-6; 43,101916-45-4; 44, 101916-46-5; 45, 74131-76-3; 46,87116-72-1; 47, 101916-47-6; 48, 101932-60-9; 49, 101916-48-7; NaSCH3, 5188-07-8; NaSC2H5, 811-51-8; NaSC,HI3, 22487-02-1; CH3SH, 74-93-1; acetyl chloride, 75-36-5; ethyl mercaptan, 75-08-1; propionyl chloride, 79-03-8.

(51) Deceased.

Organometallic Derivatives of Hormonal Steroids: 500-MHz One- and Two-Dimensional NMR Spectra of

17a-Propynylestra-l,3,5( lO)-triene-3,17P-diol and Its C O ~ ( C O ) ~ and (C5H5)2M02(C0)4 Complexes

Monique Savignac,t GBrard Jaouen,*t Charles A. Rodger,* Richard E. Perrier,§ Brian G . Sayer,§ and Michael J. McGlinchey*§

Ecole Nationale Superieure de Chimie, 75231 Paris Cedex, France, Bruker Spectrospin Ltd., Milton, Ontario, Canada L9T 1 Y6, and Department of Chemistry, McMaster University, Hamilton, Ontario, Canada L8S 4M1

Received October 16, 1985

Treatment of estrone with a propynyl Grignard reagent gives exclusively 17a-propynylestra-l,3,5(lO)-triene-3,17P-diol. This l7a-alkynyl steroid reacts with CO~(CO)~ or (C5H5)2M02(C0)4 to yield the cluster complexes (RC=CR')M2, where R = methyl, R' is the steroidal moiety, and M = CO(CO)~ or (C5H5)Mo(CO),. The cobalt complex of mestranol has likewise been prepared. The 500-MHz 'H NMR spectra of these molecules are reported and are assigned by the two-dimensional COSY technique. The shifts of the 12a- and 14a-protons of the steroid are discussed in terms of the anisotropy in diamagnetic susceptibility of the alkyne linkage, I3C spectra are also reported.

Introduction The incorporation of organometallic moieties into bio-

logically important molecules is a field of burgeoning importance. Typically, in steroid chemistry, Fe(CO), fragments may be used as temporary protecting agents,] and allylpalladium2 or Cr(C0)3 units3 have been exploited for synthetic purposes. Recently, advances in bioorgano- metallic chemistry have been directed toward immunology: and we have described the use of steroidal hormones labeled with metal carbonyls to assay receptor sites? This latter concept takes advantage of the strong infrared absorptions of metal carbonyls in the range 2100-1850

Ecole Nationale Supgrieure de Chimie.

McMaster University. f Bruker Spectroscopin Ltd.

cm-l-a window in which proteins do not absorb. Our goal is to monitor the hormone dependence of breast cancer while avoiding the use of radioactivity and its associated inconveniences. In the particular case of estrogenic hormones, it has been

reported that the 7a-, l lp-, and l7a-positions of estradiol (I) can tolerate substitution by bulky groups and still ex-

(1) Barton, D. H. R.; Gunatilika, A. A. L.; Nakanishi, T.; Patin, H.; Widdowson, D. A.; Worth, B. R. J. Chem. Soc., Perkin Trans. 1 1976,821.

(2) Trost, B. M.; Verhoeven, T. J. Am. Chem. SOC. 1978, 100, 3435. (3) (a) Top, S.; VessiBres, A,; Abjean, J.-P.; Jaouen, G. J . Chem. SOC.,

Chem. Commun. 1984,428. (b) Jaouen, G.; Top, S.; Laconi, A.; Couturier, D.: Brocard, J. J. Am. Chem. SOC. 1984, 106, 2207.

.(4) Cais, M. Actual. Chim. 1979, 7, 14. (5) (a) Jaouen, G.; Vessisres, A.; Top, S.; Ismail, A. A.; Butler, I. S. C.

R . Seances Acad. Sci., Ser. 2 1984,298, 683. (b) Jaouen, G.; VessiBres, A.; Top, S.; Ismail, A. A.; Butler, I. S. J . Am. Chem. SOC. 1985, 107, 4778.

0022-3263/86/1951-2328$01.50/0 0 1986 American Chemical Society

Eur Respir J 1997; 10: 1496–1499DOI: 10.1183/09031936.97.10071496Printed in UK - all rights reserved

Copyright ERS Journals Ltd 1997European Respiratory Journal

ISSN 0903 - 1936

Distribution of inhaled fluticasone propionate betweenhuman lung tissue and serum in vivo

N. Esmailpour*, P. Högger*, K.F. Rabe**, U. Heitmann**,M. Nakashima**, P. Rohdewald*

Distribution of inhaled fluticasone propionate between human lung tissue and serumin vivo. N. Esmailpour, P. Högger, K.F. Rabe, U. Heitmann, M. Nakashima, P. Rohdewald.©ERS Journals Ltd 1997.ABSTRACT: High retention of inhaled glucocorticoids in the airways means pro-longed anti-inflammatory action and low delivery into the serum. The objective ofthis study was to investigate the retention in and distribution of inhaled fluticas-one propionate (FP) between central and peripheral human lung tissue and serumin vivo.

In 17 patients undergoing lung resection surgery, a single 1.0 mg dose of FP wasinhaled at varying time-points (range 2.8–21.7 h) preoperatively. Peripheral andcentral lung tissue was obtained, and blood was drawn simultaneously.

FP concentrations in central lung tissue were approximately three to four timeshigher than peripheral lung tissue concentrations, which in turn, exceeded thosefound in serum by 10 times. FP was detectable up to 21 and 16 h, respectively,after inhalation, with drug levels falling almost in parallel in peripheral lung tis-sue and in serum.

The results of this study demonstrate that fluticasone propionate is retained inlung tissue for a long time. Serum concentrations after a single inhaled dose arelow. Retention of high concentrations of fluticasone propionate in the airways maypromote high topical anti-inflammatory activity.Eur Respir J 1997; 10: 1496–1499.

*Institut für Pharmazeutische Chemie,Westfälische Wilhelms-Universität Münster,Münster, Germany. **Krankenhaus Groß-hansdorf, Zentrum für Pneumologie undThoraxchirurgie, LVA Hamburg, Germany.

Correspondence: P. Rohdewald Institut für Pharmazeutische Chemie Westfälische Wilhelms-Universität 48149 Münster Germany

Keywords: Fluticasone propionateglucocorticoidinhalationplasma concentrationtissue concentration in vivo

Received: July 30 1996Accepted after revision April 20 1997

The investigation was supported by Glaxo-Wellcome.

The ideal corticosteroid for use by inhalation in thetreatment of asthma should act effectively in the airwaysand produce a minimum of systemic effects within itsdose range [1]. To achieve this, the compound shouldhave a high intrinsic topical activity (anti-inflammatorypotency) combined with low oral systemic bioavaila-bility [2]. It is favourable for a corticosteroid to displayhigh retention in the airways and low delivery into plas-ma after inhalation. Therefore, it is of interest to inves-tigate the distribution of glucocorticoids between lungtissue and blood.

Potent glucocorticoids have a high affinity for the glu-cocorticoid receptor. Of all corticosteroids tested, fluti-casone propionate (FP) has the highest in vitro affinityfor the glucocorticoid receptor in human lung [3, 4].Binding and retention studies in human lung in vitrohave demonstrated high tissue concentrations for FP(similar to beclomethasone dipropionate and its activemetabolite, but greater than budesonide, flunisolide andhydrocortisone), compared with blood plasma concen-trations [5–7].

To demonstrate the clinical relevance of these data,we present the first investigation of the tissue-plas-ma distribution of FP in humans. In this study, thedistribution of a single 1.0 mg dose of FP betweenhuman lung tissue and blood plasma was studied invivo.

Materials and methods

Patients and study design

The study was approved by the local Ethics Committee.Informed consent to participate in the study was obtainedfrom 17 patients (16 males and 1 female) with bronchialcarcinoma, who were referred for pneumectomy or loberesection. Their mean (±SD) age was 58±6 yrs (range48–67 yrs). Squamous cell carcinomas were found in11 patients, five patients had adenocarcinoma, and onepatient had a carcinoid tumour. Vital capacity ranged2.7–6.1 L, and forced expiratory volume in one second(FEV1) 58–115% of predicted (table 1).

At varying times before surgery (range 2.8–21.6 h),1.0 mg FP was administered via a large volume spacer,Volumatic™ (four puffs each of 250 µg), and the timeof administration was recorded. Thirty to 45 min be-fore intubation, 0.5 mg atropine sulphate and an anal-gesic were given to patients as premedication. Peripheraland central lung tissue free of tumour was obtained foranalysis of FP. A blood sample was drawn intraopera-tively at the time-point of ligation of the pulmonaryartery for analysis. Whole blood was centrifuged andserum samples were frozen and stored until analysed at-20°C. The time after inhalation, identical with time

N. ESMAILPOUR ET AL.

between inhalation and resection, has been defined asthe time between inhalation and the ligation of the pul-monary artery.

Materials

Chemicals were purchased from Sigma (Deisenhofen,Germany) or Merck (Darmstadt, Germany), activatedcharcoal Norit Gsx was from BDH (Wesel, Germany).Fluticasone-17-propionate (FP), tritiated FP (specific ac-tivity 55.4 Ci·mmol-1) and FP antiserum [8] were a gen-erous gift from Glaxo Wellcome (Greenford, UK). Solidphase extraction cartridges, C18-SPE-columns, were ob-tained from J.T. Baker (Philipsburg, USA), scintillationcocktail Aquasafe 500 Plus and 4 mL polyethylene vialswere from Zinsser (Frankfurt, Germany), and 2 mL incu-bation tubes were obtained from Eppendorf (Hamburg,Germany).

Assay buffer (pH 7.4 at 4°C) consisted of 100 mMtricine, 150 mM sodium chloride, and 15 mM sodiumazide. Tricine-ethanol buffer contained 5% (v/v) ethanolin assay buffer, and tricine-gelatine buffer contained0.1% gelatine (w/v) in assay buffer. Dextran-coated char-coal suspension consisted of 0.5% (w/v) Norit Gsx and0.05% (w/v) dextran. The working dilution of [3H]-FPwas 532.4 pmol·L-1, and the working solution of FPantiserum resulted in 50% binding of [3H]-FP.

Quantification of FP in lung tissue

Tissue was pulverized and one part tissue was homog-enized (Ultra-turrax, Jahnke Kunkel, Germany; PotterS, Braun Melsungen, Germany) with three parts of assaybuffer. An aliquot (2.0 g) of the tissue suspension wasfrozen in liquid nitrogen and lyophilized. The pellet wasextracted twice with 5.0 mL methanol, and an aliquotof the extract was evaporated at 50°C under a gentlenitrogen flow. Dry extracts were reconstituted in 0.1 mL

tricine-ethanol buffer and 0.1 mL tricine-gelatine buffer.Samples were mixed with 0.1 mL of FP antiserum solu-tion and incubated at 0°C for 30 min. For determinationof nonspecific binding, antiserum solution was replacedby assay buffer. After 30 min, 0.1 mL of [3H]-FP wasadded and incubation was continued for another 30 minat 0°C. The reaction was stopped by addition of 1.0 mLdextran-coated charcoal suspension. Samples were incub-ated for 10 min on ice and centrifuged. A 1.0 mL aliquotof the supernatant was counted for radioactivity. Mean(±SD) recovery was 90±3% (n=8) for tissue samples.Calibration curves were prepared with blank tissue (swinelungs). Specific binding was 75±4% for human lung tis-sue (n=8) and 80±2% for swine tissue (n=8). Thus, spe-cific binding was not statistically significantly differentbetween human and swine lung tissue. There was nodetectable difference in calibration curves obtained withswine tissue compared to calibration curves obtainedwith blank human lung tissue. Curves were linear, thecoefficient of correlation was r=0.99±0.01 (n=5). Thelower limit of quantification was 1 pg·mg-1.

Quantification of FP in serum

Frozen serum samples were thawed. Solid phase ex-traction cartridges were equilibrated with 1 mL methanol,and subsequently with 1 mL assay buffer. A 1.0 mLaliquot of serum samples was applied to the column.The cartridges were washed with 1 mL assay buffer and1 mL distilled water, and dried for 2 min under vacu-um. FP was eluted with 1.0 mL methanol, and the elu-ate was evaporated at 50°C under a gentle nitrogen flow.Radioimmunoassay (RIA) was performed as describedfor tissue samples. Recovery was 95±0.8% (n=8) forserum samples. Calibration curves were prepared withblank human serum. Curves were linear and the coeffi-cient of correlation was r=0.99±0.01 (n=6). The lowerlimit of quantification was 25 pg·mL-1.

1497

Table 1. – Patient characteristics, lung function parameters and medical data

Pt Age Sex Smoking Type of Histology VC FEV1 Time between CommentsNo. yrs status operation L % pred inhalation and

resection h

1 60 M Smoker PN AC 3.6 80 4.7 -2 61 M Smoker PN AC 3.8 80 3.7 -3 67 M Ex PN AC 3.3 79 2.8 -4 50 F Smoker LR Carcinoid 2.7 61 14.0 Plasma only5 51 M Ex PN SC 5.2 104 3.3 -6 59 M Ex PN SC 4.7 94 12.3 -7 61 M Smoker LR AC 4.1 87 3.7 -8 62 M Smoker PN SC 4.4 93 12.7 Viscous mucus9 60 M Smoker PN SC 6.1 105 13.3 -

10 57 M Smoker PN SC 3.2 58 17.5 -11 50 M Smoker PN SC 4.1 101 11.1 Viscous mucus12 56 M Ex PN SC 5.6 98 12.1 Viscous mucus13 64 M Ex PN SC 5.5 115 21.2 -14 51 M Smoker PN AC 4.2 70 16.3 -15 58 M Smoker PN SC 5.1 83 17.2 -16 48 M Smoker PN SC 3.5 67 13.3 -17 66 M Smoker PN SC 4.1 68 16.9 -

Pt: patient; M: male; F: female; Ex: ex-smoker; PN: pneumonectomy; LR: lobe resection; AC: adenocarcinoma; SC: squamouscell carcinoma; VC: vital capacity; FEV1: forced expiratory volume in one second; % pred: percentage of predicted value. FEV1% pred for males: FEV1/(4.3 × (height/100))-(0.029 × age)-2.49) × 100. FEV1 % pred for females: FEV1/(3.95 × (height/100))-(0.025 × age)-2.6) × 100.

FLUTICASONE PROPIONATE IN LUNG TISSUE

Results

FP concentrations of 0–6 ng·g-1 (0–12 nmol·kg-1) foundin peripheral lung tissue, were approximately 100 timesgreater than the FP concentrations of 0–0.11 ng·mL-1

(0–0.22 nmol·L-1) in serum, (figs. 1 and 2). Levels ofFP in peripheral lung and serum fell almost in parallelwith time after inhalation. Regression analysis of thetwo sets of data revealed high coefficients of correla-tion between concentration and time: r=0.70, p<0.01;and r=0.79, p<0.01, respectively.

The serum concentration of FP was generally below0.1 ng·mL-1, with the exception of one data point con-tributed by a patient with viscous airway mucus (thisoutlier was not used for regression analysis). Peripherallung tissue concentrations of FP could be detected upto 16.3 h, and serum concentrations up to 13.3 h afterinhalation. Concentrations of FP were generally higher(by three to four times) in central tissue, compared withconcentrations in peripheral lung tissue (fig. 1), although

interindividual variation in central lung tissue drug con-centrations was high (range <1 ng·g-1 to 23 ng·g-1).Concentrations of FP in central lung tissue persisted ata high level up to 17 h, with the exception of one pati-ent.

Discussion

This study showed that FP concentrations in centrallung tissue were about three to four times higher thanin peripheral lung tissue, which in turn exceeded thosefound in serum by 100 times. No relevant informationcan be drawn from serum levels of inhaled steroids aboutthe concentrations in the target organ. The present inves-tigation gives more useful information than pharmaco-kinetic studies analysing serum levels only, because drugconcentrations in the lung became directly available.The study design allowed the time course of drug con-centrations to be described by combining single datafrom a sufficiently high number of patients.

Human in vivo data for the tissue-serum distributionof inhaled glucocorticoids and drug deposition in cen-tral and peripheral lung has not previously been inves-tigated. Only the distribution of a glucocorticoid betweenthe peripheral lung tissue and plasma of patients wasstudied after inhalation of 1.6 mg budesonide [9].

The higher FP concentrations in central lung tissuecompared with peripheral lung tissue is not unexpect-ed, since deposition of an inhaled drug is likely todecrease with greater distance from the trachea. Thismay also explain differences in drug deposition foundin central lung tissue between patients where tissue sam-ples were obtained from bronchi of different diameter.No close correlation was found for FP concentrationsin the central airways and FEV1, vital capacity or thepresence of mucus. However, in two of the three patientshaving viscous mucus in the upper airways, we foundhigh FP concentrations of about 20 ng·g-1, and anotherpatient with 22 ng·g-1 FP in central lung tissue presentedthe lowest FEV1.

Systematic investigations of drug deposition and dis-tribution to tissue in the central airways are needed, toobtain a better understanding of drug deposition as afunction of the distance from the trachea as well as afunction of airway diameter. The higher FP concentra-tions in central lung tissue persisted for more than 17h, in contrast to the steadily declining concentrations inperipheral lung tissue. There are two probable reasonsfor this difference. Firstly, FP particles deposited in cen-tral lung tissue are likely to be of bigger mean size thanthose deposited in peripheral lung tissue. Secondly, thedensity of blood capillaries is higher in peripheral thanin central tissue, so that the tissue-plasma exchange rateis higher in peripheral tissue and FP tissue concentra-tions decrease faster.

The lung tissue concentration of FP found in this studyis estimated to be approximately 10–100 times higherthan the concentration of FP previously calculated tooccupy 50% of human glucocorticoid receptor in vitro[4]. This means that the tissue receptor binding sitesshould be 100% saturated with FP, still leaving a non-(receptor)-bound excess. Thus, FP can exhibit its anti-inflammatory action over a longer period of time, since

1498

◆

◆

◆◆◆◆◆◆

◆

◆◆◆

◆

◆◆◆

0

5

10

15

20

25

Tiss

ue c

once

ntra

tion

ng·

g-1

0 5 10 15 20Time after inhalation h

●

●

● ●●

●●

● ●● ● ●●●● ●

●

0.12

0.10

0.08

0.06

0.04

0.02

00 5 10 15 20

Time after inhalation h

Plas

ma

conc

entra

tion

ng·

mL-1

Fig. 1. – Concentration of fluticasone propionate (FP) in central andperipheral lung tissue after inhalation of 1 mg. Symbols represent sin-gle patient data. : FP concentration in central lung tissue; ◆: FP con-centration in peripheral lung tissue

Fig. 2. – Concentration of fluticasone propionate (FP) in humanserum after inhalation of 1 mg. Symbols represent single patient data.●: FP concentration in serum.

Seru

m c

once

ntra

tion

ng·

mL-

1

christinehurd

Highlight

christinehurd

Highlight

christinehurd

Highlight

it is stored in lung tissue for a prolonged period in suf-ficiently high concentrations. The prolonged presenceof FP in lung tissue, together with the high relativereceptor affinity to the glucocorticoid receptor, helps toexplain the high clinical efficacy of FP.

Under steady-state conditions it is to be expected thatFP concentrations in lung tissue and plasma are higherthan in the present study after a single dose, becauseplasma concentrations of FP under steady-state condi-tions are about 50% higher compared to single dosing[10].

In the present study, the FP concentrations in periph-eral lung tissue following inhalation of a 1.0 mg dosewere twice as high as those found for budesonide in asimilar study [9]. Whereas linear regression analysisshowed a statistically significant correlation between FPconcentration and time in peripheral lung tissue (r=0.74;p>0.001), a significant correlation was not demonstra-ble with budesonide concentrations in lung parenchyma(r=0.37; p=NS). This might be due to high variation ofdata, or to the fact that lung tissue and serum concen-trations were measured up to 4 h only. In contrast totissue concentrations, VAN DEN BOSCH et al. [9] foundthat budesonide concentrations in plasma were consid-erably higher than FP serum concentrations.

The results of the present study and of the study withbudesonide [9] confirm our in vitro data with humanlung tissue [5–7]. After saturation of human lung tissuein vitro with glucocorticoids and subsequent incubationin serum, the total concentration of FP remaining in thehuman lung tissue was about three times higher than forbudesonide after 1 h in vitro, whereas FP concentra-tions in vivo were twice as high as for budesonide.

The parallel decline of FP concentrations in periph-eral lung tissue and serum indicates that the delivery ofFP from tissue to serum is the rate-limiting step for theelimination of FP.

Whereas the correlation of in vitro with in vivo datawas high if the study was performed with human lungtissue [5–7], in vitro data obtained with rat tissue [10]did not correspond with the results of this FP in vivostudy or with the budesonide in vivo study [9], sinceMILLER-LARSSON et al. [11] found that 40–50% of budes-onide and only 30–40% of FP was left in the tracheain vitro 2 h after administration of the glucocorticoid.

The preclinical results that we obtained with flutica-sone propionate, such as its affinity to the human glu-cocorticoid receptor [3], the receptor kinetics [4], and

the binding to and retention in human lung tissue invitro [5–7] concurrently proved that fluticasone propi-onate is a long-acting and very potent glucocorticoid.The in vivo study presented here is in good accordancewith these previous findings.

References

1. Ryrfeldt Å, Andersson P, Edsbacker S, Tonnesson M,Davies D, Pauwels R. Pharmacokinetics and metabo-lism of budesonide, a selective glucocorticoid. Eur JRespir Dis 1982; 63 (Suppl. 122): 86–95.

2. Pauwels R. Use of corticosteroids in asthma. In: D'ArcyPF, McElnay JL, eds. Pharmacy and Pharmacotherapyof Asthma. Chichester, Ellis Harwood, 1989; pp. 104–118.

3. Würthwein G, Rehder S, Rohdewald P. Lipophilicityand receptor affinity of glucocorticoids. Pharm Zgt Wiss1992; 137: 161–167.

4. Högger P, Rohdewald P. Binding kinetics of flutica-sone propionate to the human glucocorticoid receptor.Steroids 1994; 59: 597–602.

5. Högger P, Bonsmann U, Rohdewald P. Distribution andintrinsic activity of inhaled glucocorticoids in vitro. EurJ Pharm Sci 1994; 2: 159.

6. Högger P, Bonsmann U, Rohdewald P. Efflux of glu-cocorticoids from human lung tissue to human plasmain vitro. Eur Respir J 1994; 7 (Suppl. 18): 382s.

7. Rohdewald P, Bonsmann U, Högger P. Die Bindunginhalativer Glukokortikoide an menschliches Lungen-gewebe in vitro. In: Leupold W, Nolte D, eds. NeueAspekte der Inhalativen Glukokortikoid-Therapie desAsthma Bronchiale. München-Deisenhofen, Dustri-VerlagDr. K. Feistle. 1995; pp. 14–27.

8. Bain BM, Harrison G, Jenkins KD, Pateman AJ, ShenoyEVB. A sensitive radioimmunoassay, incorporating solid-phase extraction, for fluticasone-17-propionate in plas-ma. J Pharm Biomed Anal 1993; 11: 557–561.

9. Van den Bosch JMM, Westermann CJJ, Aumann J,Edsbäcker S, Tönnisson M, Selroos O. Relationshipbetween lung tissue and blood plasma concentrations ofinhaled budesonide. Biopharm Drug Disp 1993; 14: 455-459.

10. Thorsson L, Källén A, Wirén J-E, Paulson J. Pharmaco-kinetics of inhaled fluticasone propionate. Eur Respir J1996; (Suppl. 23): 1645.

11. Miller-Larsson A, Mattsson H, Ohlsson S, et al. Prolongedrelease from the airway tissue of glucocorticoids budes-onide and fluticasone propionate as compared to beclo-methasone dipropionate and hydrocortisone. Am J RespirMed 1994; 149: A466.

N. ESMAILPOUR ET AL. 1499

ACTH = adrenocorticotropic hormone; AP-1 = activator protein-1; CBP = CREB-binding protein; COX = cyclooxygenase; CREB = cAMP-response-element-binding protein; ERK = extracellularly regulated kinase; GR = glucocorticoid receptor; GRE = glucocorticoid response element; IL = interleukin;JNK = c-Jun N-terminal kinase; MAP = mitogen-activated protein; MKK = MAP kinase kinase; MKP = MAP kinase phosphatase; NF = nuclear factor.

Available online http://arthritis-research.com/content/4/3/146

IntroductionThe anti-inflammatory action of glucocorticoid hormoneswas discovered by Hench and colleagues over 50 yearsago [1]. Hench had noted that the symptoms of rheuma-toid arthritis were often improved in pregnancy and whena patient had jaundice, both situations in which, he rea-soned, there was an increase in steroids in the body. Theactive substances of the adrenal cortex had then recentlybeen isolated by Reichstein and Kendall and shown to besteroids. It seemed possible that these might alleviateinflammatory symptoms. Hench and his co-workers foundthat small doses of cortisone dramatically improved thesymptoms of patients with rheumatoid arthritis. Hench,

Kendall, and Reichstein were jointly awarded the NobelPrize in physiology and medicine in 1950. Powerful syn-thetic glucocorticoids were then developed, which,despite their unwelcome side effects, became and remainmainstays of anti-inflammatory and immunosuppressivetherapy.

The side effects of glucocorticoids that severely limitedtheir use were osteoporosis, diabetes, hypertension,cataracts, thinning of the skin, and the characteristicappearance of Cushing’s syndrome. They also sup-pressed the hypothalamic–pituitary–adrenal axis andarrested growth. Their potency in suppressing inflamma-

CommentaryGlucocorticoids: do we know how they work?Jeremy Saklatvala

Kennedy Institute of Rheumatology Division, Faculty of Medicine, Imperial College of Science, Technology and Medicine, London, UK

Correspondence: Jeremy Saklatvala, Kennedy Institute of Rheumatology Division, Faculty of Medicine, Imperial College of Science, Technology andMedicine, London W6 8LH, UK. Tel: +44 (0) 20 8383 4444; fax: +44 (0) 20 8563 0399; e-mail [email protected]

Abstract

It is not known to what extent glucocorticoid hormones cause their anti-inflammatory actions and theirundesirable side effects by the same or different molecular mechanisms. Glucocorticoids combinewith a cytoplasmic receptor that alters gene expression in two ways. One way is dependent on thereceptor’s binding directly to DNA and acting (positively or negatively) as a transcription factor. Theother is dependent on its binding to and interfering with other transcription factors. Both mechanismscould underlie suppression of inflammation. The liganded receptor binds and inhibits the inflammatorytranscription factors activator protein-1 and NF-κB. It also directly induces anti-inflammatory genessuch as that encoding the protein inhibitor of NF-κB. Recent work has shown that glucocorticoidsinhibit signalling in the mitogen-activated protein kinase pathways that mediate the expression ofinflammatory genes. This inhibition is dependent on de novo gene expression. It is important toestablish the significance of these different mechanisms for the various physiological effects ofglucocorticoids, because it may be possible to produce steroid-related drugs that selectively targetthe inflammatory process.

Keywords: glucocorticoid, inflammation, JNK, MAP kinase, NF-κB

Received: 24 October 2001

Revisions requested: 8 November 2001

Revisions received: 22 November 2001

Accepted: 26 November 2001

Published: 21 January 2002

Arthritis Res 2002, 4:146-150

This article may contain supplementary data which can only be foundonline at http://arthritis-research.com/content/4/3/146

© 2002 BioMed Central Ltd(Print ISSN 1465-9905; Online ISSN 1465-9913)


tion stimulated much investigation into their mechanism ofaction. Glucocorticoids inhibit expression of many of thegenes involved in inflammatory and immune responses.These include those encoding cytokines, chemokines,cell-surface receptors, adhesion molecules, tissue factor,degradative proteinases, and enzymes such as cyclo-oxygenase (COX)-2 and inducible nitric oxide synthase,which produce inflammatory mediators.

The glucocorticoid receptor (GR) [2] binds to specificDNA sequences – glucocorticoid response elements(GREs) – which may be positive or negative, either activat-ing or repressing transcription, respectively (Fig. 1). In theabsence of ligand, the receptor is cytoplasmic and is com-plexed with chaperones, including hsp90, hsp56(immunophilin), and calreticulin. Upon ligand binding tothe receptor, the chaperones are shed, exposing nuclearlocalisation signals. The receptor dimerises when it bindsto a GRE. Genes that are positively regulated by GREsinclude those involved in gluconeogenesis, such as thosefor tyrosine aminotransferase, alanine aminotransferase,and phosphoenolpyruvate carboxykinase. Examples ofgenes negatively regulated by GREs are those for pro-opi-omelanocortin (the ACTH precursor) and prolactin. Thegenes responsible for such side effects of glucocorticoidsas osteoporosis, diabetes, and hypertension are largelyunknown. Understanding the transcriptional basis of thedifferent physiological effects is important because it maybe possible to develop more selective agents, whichtarget the inflammatory process.

Glucocorticoid receptors interact withinflammatory transcription factorsNo genes responding positively to glucocorticoid hor-mones have been identified that explain the profound sup-pression of the inflammatory response. Furthermore,genes of the inflammatory response lack repressingGREs, so other explanations for the suppression of inflam-mation have been sought [3,4]. Currently, it is thought thata major mechanism is transcriptional interference (seeFig. 1). Inflammatory-response genes are typically regu-lated by the transcription factors NF-κB and activatorprotein-1 (AP-1). The liganded GR interacts with AP-1complexes (generally c-Jun/c-Fos heterodimers) and pre-vents their transcriptional activity [3,4]. Similar tetheringoccurs with NF-κB [5,6]. This interference between tran-scription factors is reciprocal: not only can GR prevent thefunction of AP-1 and NF-κB, but also AP-1 and NF-κB canprevent transcriptional activation by GR. The physiologicalsignificance of these mechanisms is still speculative andthe interactions between the endogenous proteins havebeen difficult to demonstrate in vivo. Another possibleexplanation for the functional competition between AP-1,NF-κB, and GR is that they compete for the transcriptionalcoactivators CREB-binding protein (CBP) and p300 [7,8].However, this notion is controversial.

An additional mechanism of transcriptional interferencehas been recently proposed [9]. Corticosteroid interfereswith IL-1-induced gene activation by inhibiting the histoneacetylation that loosens chromatin structure, and enables

Figure 1

Mechanisms of transcriptional regulation by glucocorticoids (adapted from M Karin [3]). Liganded GRs (ovals with patches) regulate transcriptionby direct binding to DNA elements (GREs) or by binding to other transcription factors (tethering). GRE-regulated and NF-κB/AP-1-regulatedpromoters are shown schematically with single binding sites; typically, there may be one or several sites. Interference with the activity of AP-1(open and dark ovals) or NF-κB (open and dark squares) after tethering could be a direct effect on the factors’ function, or it could be aninteraction with coactivators or components of the TIC. AP-1 = activator protein-1; GR = glucocorticoid receptor; GRE = glucocorticoid responseelement; NF-κB = nuclear factor-κB; TIC = transcriptional initiation complex.

Arthritis Research Vol 4 No 3 Saklatvala

access of transcription factors to their DNA binding sites[9]. The liganded GR bound to the transcriptional com-plexes inhibits their acetyltransferase activity. This couldbe an adjunct to the mechanism of tethering of transcrip-tion factors.

Glucocorticoids may also interfere with NF-κB activationby a mechanism dependent upon de novo gene expres-sion. NF-κB is sequestered in the cytoplasm with aninhibitor, IκBα, whose degradation is induced by inflam-matory stimuli. Production of this inhibitor is increased bydexamethasone [10,11]. This effect is slow and its physio-logical significance remains to be established.

Dexamethasone interferes with pathways ofmitogen-activated protein kinaseRecently, another possible mechanism for glucocorticoidaction has become apparent: dexamethasone inhibits sig-nalling in mitogen-activated protein kinase (MAP kinase)pathways, which are activated by inflammatory stimuli [12].This suggests that glucocorticoids may block inflammatorysignalling at a level above transcription factor activation.The effect, unlike transcriptional interference, requires geneinduction, possibly of a MAP kinase phosphatase (MKP),and could inhibit both transcriptional and post-transcrip-tional mechanisms controlled by the MAP kinases (Fig. 2).

There are three types of MAP kinase [13] and they partici-pate in distinct phosphorylation cascades (Fig. 3). Theprototype MAP kinase is the extracellularly regulatedkinase (ERK), or p42, which is activated by many stimuli.The other two, the c-Jun N-terminal kinase (JNK) and p38,are very strongly activated by inflammatory stimuli such asIL-1, tumour necrosis factor, and microbial products (e.g.lipopolysaccharide), and by cellular stress (e.g. UV). TheMAP kinases are activated by their specific MAP kinasekinases (MKKs) by dual phosphorylation of a motif com-prising a threonyl and a tyrosyl residue separated by asingle amino acid. The MAP kinases phosphorylate varioussubstrates, including other protein kinases. At the bottomof the cascades lie proteins controlling gene expressionand other processes.

Serving to modulate the activity of MAP kinases and toswitch off the pathways after the response to a stimulus,are MKPs [14]. These dual-specificity phosphatasesremove phosphates from both the threonyl and tyrosylresidues in the MAP kinase activation motifs. More than12 of these phosphatases are known to exist, differing inexpression, subcellular localisation, and specificity for thedifferent MAP kinases. Since removal of phosphate fromeither threonyl or tyrosyl residues inactivates theenzymes, they may also be regulated by phospho-

Figure 2

Glucocorticoids may act (1) by transcriptional interference and (2) by inhibiting MAP kinases. The left side represents glucocorticoid (G) combiningwith its receptor (GR) and inducing gene expression by binding to DNA glucocorticoid response elements (GREs). The right side shows aninflammatory stimulus (e.g. lipopolysaccharide, IL-1, tumour necrosis factor) activating protein kinase cascades (see Fig. 3) and inducinginflammatory response genes. (1) Transcriptional interference is due to the liganded GR directly binding the transcription factors activator protein-1and nuclear factor-κB and inhibiting their action. (2) Glucocorticoid-induced genes, possibly MKPs, inhibit MAP kinase signalling pathways bykeeping them in the dephosphorylated state. This would inhibit both transcriptional and post-transcriptional mechanisms underlying inflammatorygene expression. MAP = mitogen-activated protein; MKP = MAP kinase phosphatase.

christinehurd

Highlight

serine/phosphothreonine protein phosphatases andprotein phosphotyrosine phosphatases.

The three MAP kinase pathways, in addition to the proteinkinase system that activates NF-κB by phosphorylating theinhibitor IκB, are the major signalling systems throughwhich genes of the inflammatory response are activated.Such genes are activated by several transcription factorsand typically, as mentioned earlier, NF-κB and AP-1 arecrucial. The JNK pathway is a major route for activation ofc-Jun, and therefore of AP-1 complexes. The mRNAs ofmany cytokines and key regulators of the inflammatoryresponse are unstable by virtue of clustered AUUUAmotifs in their 3′ untranslated regions [15]. The p38 MAPkinase pathway stabilises such mRNAs [16–21]. It is pre-sumed that downstream kinases regulate proteins thatbind to the instability motifs.

The role of the p38 MAP kinase pathway has been eluci-dated by studying its effect on the function of the 3′untranslated regions of the COX-2 [20] and IL-8 mRNAs[21]. Because dexamethasone destabilises some mRNAs,including the mRNA of COX-2 [22], its mechanism wasinvestigated. It antagonised the stabilising action of the

p38 MAP kinase on the COX-2 3′ UTR. This raised thepossibility that dexamethasone was inhibiting p38 MAPkinase. There had been reports that glucocorticoids inhib-ited JNK activity in cells [23,24] and suppressed activity ofthe ERK cascade in mast cells [25]. Treating cells with dex-amethasone for 1–2 hours before stimulation inhibited theactivation of both p38 MAP kinase and JNK by UV light, byIL-1, or by bacterial lipopolysaccharide [12]. We know thatthis inhibition was at the level of the MAP kinase, becausedexamethasone had no effect on the p38 activator, MKK6,and the inhibition could be circumvented by transfecting anactive mutant of the substrate of p38, the MAP kinase-acti-vated protein kinase-2 [12]. The effect of dexamethasoneon the p38 MAP kinase was dependent upon GR and denovo mRNA synthesis.

An explanation for these findings is that steroids induce aprotein phosphatase that keeps the p38 MAP kinase (andJNK) in the dephosphorylated state. Alternatively, thesteroid could be inducing a molecule that inhibits theMKKs. It will be interesting to know the basis of the MAPkinase inhibition, its dependence on GRE-regulatedgenes, and its importance in inflammation relative to themechanism of transcriptional interference.


Figure 3

The MAP kinase cascades. This highly simplified scheme shows the three major MAP kinase pathways of mammalian cells. The MAP kinases areERK, JNK, and p38. A few downstream substrates are shown. ATF = activating transcription factor; cPLA2 = cytosolic phospholipase A2; CREB =cAMP-response-element-binding protein; ERK = extracellularly regulated kinase; JNK = c-Jun N-terminal kinase; LPS = lipopolysaccharide; MAP = mitogen-activated protein; MAPKAPK = MAP kinase-activated protein kinase; MKK = MAP kinase kinase; MKKK = MAP kinase kinasekinase; TNF = tumour necrosis factor.

Arthritis Research Vol 4 No 3 Saklatvala

Note added in proofGlucocorticoid increasing expression of MKP-1 hasrecently been reported by Kassel et al. [26].

AcknowledgementsThe author is grateful to Mrs Suzie Johns for preparation of the manu-script, to the many authors whose work has not been directly cited inthis short review, and to the Arthritis Research Campaign and theMedical Research Campaign for financial support.

References1. Hench PS, Slocumb CH, Barnes AR, Smith HL, Polley HF,

Kendall EC: The effect of a hormone of the adrenal cortex, 17-hydroxy-11-dehydrocorticosterone (compound E), on theacute phase of rheumatic fevers. Proceedings of the StaffMeetings of the Mayo Clinic 1949, 24-277-297.

2. Beato M, Truss M, Chavez S: Control of transcription by steroidhormones. Ann N Y Acad Sci 1996, 784:93-123.

3. Karin M: New twists in gene regulation by glucocorticoidreceptor: is DNA binding dispensable? Cell 1998, 93:487-490.

4. Newton R: Molecular mechanisms of glucocorticoid action:what is important? Thorax 2000, 55B, 603-613.

5. Ray A, Prefontaine KE: Physical association and functionalantagonism between the p65 subunit of transcription factorNF-kappa B and the glucocorticoid receptor. Proc Natl AcadSci U S A 1994, 91:752-756.

6. Scheinman RI, Gualberto A, Jewell CM, Cidlowski JA, Baldwin ASJr: Characterization of mechanisms involved in transrepres-sion of NF-kappa B by activated glucocorticoid receptors. MolCell Biol 1995,15:943-953.

7. Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B, Lin SC,Heyman RA, Rose DW, Glass CK, Rosenfeld MG: A CBP inte-grator complex mediates transcriptional activation and AP-1inhibition by nuclear receptors. Cell 1996, 85:403-414.

8. Sheppard KA, Phelps KM, Williams AJ, Thanos D, Glass CK,Rosenfeld MG, Gerritsen ME, Collins T: Nuclear integration ofglucocorticoid receptor and nuclear factor-kappaB signalingby CREB-binding protein and steroid receptor coactivator-1. JBiol Chem 1998, 273:29291-29294.

9. Ito K, Barnes PJ, Adcock IM: Glucocorticoid receptor recruit-ment of histone deacetylase 2 inhibits interleukin-1beta-induced histone H4 acetylation on lysines 8 and 12. Mol CellBiol 2000, 20:6891-6903.

10. Scheinman RI, Cogswell PC, Lofquist AK, Baldwin AS Jr: Role oftranscriptional activation of I kappa B alpha in mediation ofimmunosuppression by glucocorticoids. Science 1995,270:283-286.

11. Auphan N, DiDonato JA, Rosette C, Helmberg A, Karin M:Immunosuppression by glucocorticoids: inhibition of NF-kappa B activity through induction of I kappa B synthesis.Science 1995, 270:286-290.

12. Lasa M, Brook M, Saklatvala J, Clark AR: Dexamethasonedestabilizes cyclooxygenase 2 mRNA by inhibiting mitogen-activated protein kinase p38. Mol Cell Biol 2001, 21:771-80.

13. Garrington TP, Johnson GL: Organization and regulation ofmitogen-activated protein kinase signaling pathways. CurrOpin Cell Biol 1999, 11:211-218.

14. Keyse SM: Protein phosphatases and the regulation ofmitogen-activated protein kinase signalling. Curr Opin CellBiol 2000, 12:186-92.

15. Shaw G, Kamen R: A conserved AU sequence from the 3′′untranslated region of GM-CSF mRNA mediates selectivemRNA degradation. Cell 1986, 46:659-667.

16. Ridley SH, Dean JL, Sarsfield SJ, Brook M, Clark AR, Saklatvala J:A p38 MAP kinase inhibitor regulates stability of interleukin-1-induced cyclooxygenase-2 mRNA. FEBS Lett 1998, 439:75-80.

17. Dean JL, Brook M, Clark AR, Saklatvala J: p38 mitogen-activatedprotein kinase regulates cyclooxygenase-2 mRNA stabilityand transcription in lipopolysaccharide-treated human mono-cytes. J Biol Chem 1999, 274:264-269.

18. Miyazawa K, Mori A, Miyata H, Akahane M, Ajisawa Y, Okudaira H:Regulation of interleukin-1beta-induced interleukin-6 geneexpression in human fibroblast-like synoviocytes by p38mitogen-activated protein kinase. J Biol Chem 1998, 273:24832-24838.

19. Brook M, Sully G, Clark AR, Saklatvala J: Regulation of tumournecrosis factor alpha mRNA stability by the mitogen-activatedprotein kinase p38 signalling cascade. FEBS Lett 2000, 483:57-61.

20. Lasa M, Mahtani KR, Finch A, Brewer G, Saklatvala J, Clark AR:Regulation of cyclooxygenase 2 mRNA stability by themitogen-activated protein kinase p38 signaling cascade. MolCell Biol 2000, 20:4265-4274.

21. Winzen R, Kracht M, Ritter B, Wilhelm A, Chen CY, Shyu AB,Muller M, Gaestel M, Resch K, Holtmann H: The p38 MAP kinasepathway signals for cytokine-induced mRNA stabilization viaMAP kinase-activated protein kinase 2 and an AU-rich region-targeted mechanism. Embo J 1999, 18:4969-4980.

22. Ristimaki A, Narko K, Hla T: Down-regulation of cytokine-induced cyclo-oxygenase-2 transcript isoforms by dexam-ethasone: evidence for post-transcriptional regulation.Biochem J 1996, 318:325-331.

23. Swantek JL, Cobb MH, Geppert TD: Jun N-terminalkinase/stress-activated protein kinase (JNK/SAPK) isrequired for lipopolysaccharide stimulation of tumor necrosisfactor alpha (TNF-alpha) translation: glucocorticoids inhibitTNF-alpha translation by blocking JNK/SAPK. Mol Cell Biol1997, 17:6274-6282.

24. Caelles C, Gonzalez-Sancho JM, Munoz A: Nuclear hormonereceptor antagonism with AP-1 by inhibition of the JNKpathway. Genes Dev 1997, 11:3351-3364.

25. Rider LG, Hirasawa N, Santini F, Beaven MA: Activation of themitogen-activated protein kinase cascade is suppressed bylow concentrations of dexamethasone in mast cells. JImmunol 1996, 157:2374-2380.

26. Kassel O, Sancono A, Kratzschmar J, Kreft B, Stassen M, Cato AC:Glucocorticoids inhibit MAP kinase via increased expressionand decreased degradation of MKP-1. EMBO J 2001,20:7108-7116.

Respiratory Medicine (2012) 106, 9e14

Available online at www.sciencedirect.com

journal homepage: www.elsevier .com/locate/rmed

REVIEW

The role of mast cells in allergic inflammation

Kawa Amin a,b,*

aDepartment of Medical Science, Respiratory Medicine and Allergology, Clinical Chemistry and Asthma Research Centre,Uppsala University and University Hospital, Uppsala, SwedenbDepartment of Medical Microbiology/Immunology, School of Medicine, Faculty of Medical Science, SulaimaniUniversity,Sulaimani, Iraq

Received 3 April 2011; accepted 20 September 2011Available online 22 November 2011

KEYWORDSAllergic asthma;Mast cells;Inflammation;Cytokines;Remodeling;B- and T-lymphocytes

Abbreviations: AA, allergic asthma; Aextracellular matrix; ELAM-1, endothemolecule; INF-g, interferon-gamma;platelet-activating factor; TNF, tumor* Present address: Department of Me

Uppsala University and University HosE-mail address: kawa.amin@medsc

0954-6111/$ - see front matter ª 201doi:10.1016/j.rmed.2011.09.007

Summary

The histochemical characteristics of human basophils and tissue mast cells were describedover a century ago by Paul Ehrlich. When mast cells are activated by an allergen that bindsto serum IgE attached to their Fc 3RI receptors, they release cytokines, eicosanoids and theirsecretory granules. Mast cells are now thought to exert critical proinflammatory functions,as well as potential immunoregulatory roles, in various immune disorders through the releaseof mediators such as histamine, leukotrienes, cytokines chemokines, and neutral proteases(chymase and tryptase). The aim of this review is to describe the role of mast cells in allergicinflammation.

Mast cells interact directly with bacteria and appear to play a vital role in host defenseagainst pathogens. Drugs, such as glucocorticoids, cyclosporine and cromolyn have been shownto have inhibitory effects on mast cell degranulation and mediator release. This review showsthat mast cells play an active role in such diverse diseases as asthma, rhinitis, middle ear infec-tion, and pulmonary fibrosis.

In conclusion, mast cells may not only contribute to the chronic airway inflammatoryresponse, remodeling and symptomatology, but they may also have a central role in the initi-ation of the allergic immune response, that is providing signals inducing IgE synthesis by B-lymphocytes and inducing Th2 lymphocyte differentiation.ª 2011 Elsevier Ltd. All rights reserved.

PC, antigen presenting cell; BAL,lial-leukocyte adhesion molecule-IL, interleukin; LTs, leukotriencenecrosis factor; VCAM-1, vasculadical Science, Respiratory Medicinpital, Uppsala, Sweden.i.uu.se.

1 Elsevier Ltd. All rights reserved

bronchoalveolar lavage; CTMC, connective tissue mast cell; ECM,1; Fc 3RI, high-affinity IgE receptors I; ICAM-1, intercellular adhesions; MC, mast cell; MMC, Mucosal mast cell; PGs, prostaglandins,r cell adhesion molecule.e and Allergology, Clinical Chemistry and Asthma Research Centre,

.


http://dx.doi.org/10.1016/j.rmed.2011.09.007

www.sciencedirect.com/science/journal/09546111

http://www.elsevier.com/locate/rmed



10 K. Amin

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10Mast cell development and differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10Mast cell activation and mediator production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10Functions of mast cells in physiological and pathological states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11Mast cells and airway remodeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

Introduction

Mast cells are found in the skin and in all mucosal tissues athomeostasis, and numbers are elevated in asthmatics lungs1

and gastrointestinal tract of inflammatory bowel disease.Mast cells were first described by Ehrlich in his 1878 doctoralthesis on thebasis of their unique staining characteristics andlarge granules, that gave them their name, “Mastzellen”which means well-fed cells, because their cytoplasm wasstuffedwith granularmaterial. Mast cells are nowconsideredtobepart of the immune system.Themast cellwas identifiedas a mesenchymal cell which is stained metachromaticallywith someblue dyes and it was recognized several years laterthat these cells contained in their granules the majority ofthe body’s histamine.2 Mast cells play a central role ininflammatory and immediate allergic reactions. They areable to release potent inflammatory mediators, such ashistamine, proteases, chemotactic factors, cytokines andmetabolites of arachidonic acid that act on the vasculature,smooth muscle, connective tissue, mucous glands andinflammatory cells.3 Histamine is not only released when thebody encounters a toxic substance, it is also released whenmast cells detect injury. It causes nearby blood vessels todilate allowing more blood to reach the site of the injury orinfection. Mast cells are localized in the connective tissueand do not usually circulate in the blood stream.

The aim of this review is to discuss the effects ofdifferent Th2 cytokines on mast cell development, and thecontribution of these cells to the chronic airway inflam-matory response, tissue remodeling and symptomatology,and also to understand the role of these cells at the initi-ation of the allergic immune response, where they providesignals inducing IgE synthesis by B-lymphocytes and Th2lymphocyte differentiation.

Mast cell development and differentiation

Mast cells arise in the bone marrow where maturation isinfluencedbystemcell factor binding to the receptor c-kit andby other cytokines such as interleukin (IL)-3, IL-4, IL-9, and IL-10. These cytokines promote differentiation and proliferationof both human and mouse mast cells.4e7 The SCF receptor (c-kit) plays an important role in the hematopoiesis duringembryonic development. Mast cell is the only terminallydifferentiated hematopoietic cell that expresses the c-Kit

receptor. In addition, SCF promotes mast cell adhesion,migration, proliferation, and survival.8 SCF also promotes therelease of histamine and tryptase, which are involved in theallergic response.Mastcell progenitors leavethebonemarrowand settle in various tissues dependent of stimulation.9 Twotypes ofmast cells, mucosal and connective tissuemast cells,were reported in rodent tissue in the 1960’s on the basis ofhistochemicalandfixationcharacteristics that reflect, inpart,whether heparin proteoglycan was present in secretorygranules. Mucosal mast cell (MMC) granules stain blue withcopper phthalocyanin dyes, such as Astra blue or Alcian blue,in a staining sequence with safranin, while connective tissuemast cell (MCTC) granules stain red.10MC(T) andMC(TC) typesof humanmast cells (MCs) are distinguished from one anotheron the basis of the protease compositions of their secretorygranules, but their structural, functional differences anddevelopmental relationships have beenwell characterized byother authors.11,12 Mast cells are long-lived, surviving formonth or even years, in the tissue. Evolutionary, mast cellsexisted and participated in host defense long before thedevelopment of cells of adaptive immune system. Increasednumbers of MCT and MCTC mast cells are seen in fibroticdiseases whereas its numbers are relatively unchanged inallergic or parasitic diseases and in HIV infection.13 Thepresence of these MCTC cells could help explain why patientswith HIV infection continue to have allergic reactions. TheMCTC mast cell, however, expresses tryptase, chymase. Ittends to predominate in the respiratory tract, gastrointestinaltract as well as in skin, synovium, and subcutaneous tissue.

Mast cell activation and mediator production

The cytoplasm of mast cells contains organelles: lipid bodieswhere metabolism of arachidonic acid occurs and where theproducts of this metabolism, including leukotrienes, arestored.14 Cytokines and histamine are other products foundin mast cells organelles (Fig. 1). These organelles are proneto exocytosis and extracellular release of mediators. Therelease may be induced by: (a) chemical substances, such astoxins, venoms, and proteases; (b) endogenous mediators,including tissue proteases, cationic proteins derived fromeosinophils and neutrophils; (c) immune mechanisms thatmay be IgE-dependent or IgE-independent. IgE-dependentdegranulation is a consequence of the preferential produc-tion of IgE, in response to certain antigens (allergens). Duringan allergic response IgE release from B-cells will bind tomast

Figure 1 The IgE-primed mast cell releases granules andpowerful chemical mediators, such as histamine, cytokines,granulocyte macrophage colony-stimulating factor (GM-CSF),leukotrienes, heparin, and many proteases into the environ-ment. These chemical mediators cause the characteristicsymptoms of allergy.

The role of mast cells in allergic inflammation 11

cells, blanketing the plasma membranes of these immunecells. Half a million IgE molecules coat the surface of mastcells, binding to the high-affinity IgE receptors (Fc 3RI) onmembranes with the Fc portion. This leaves their Fab, orantigen binding segment, free to bind the antigen.9,13,15,16 Asubsequent exposure to the same allergen cross-links thecell-bound IgE and triggers the release of preformed pros-taglandins, histamines and cytokines (Fig. 2).9,14,17e19 Mastcell degranulation is preceded by increased Ca2þ influx,which is a crucial process; ionophores that increase cyto-plasmic Ca2þ also promote degranulation, whereas agentswhich deplete cytoplasmic Ca2þ suppress degranulation(Fig. 2).18,20 Additionally, in some cases, otherligandereceptor interactions, summarized in Fig. 1, can leadto mast cell degranulation.

Newly generated mediators, often absent in the restingmast cells, are as well produced during IgE-mediated acti-vation, and consist of arachidonic acid metabolites, prin-cipally leukotriene C4 (LTC4), prostaglandin D2 (PGD2) andof cytokines.21e23 Of particular interest in humans is theproduction of tumor necrosis factor (TNF-a, b), and inter-leukin (IL)-4, IL-5, IL-6, IL-1b and IL-13.9,24e26 Those lipidmediators and cytokines and preformed histamine, canhave profound effects on vascular endothelium, includingthe alteration of vascular permeability and adhesiveness.This can allow other circulating inflammatory cells toadhere to the endothelium and to migrate into thesurrounding tissue. Cytokines and lipid mediators do as wellelicit a direct influence on lymphocytes and macrophages inthe murine system.27 IL-4, IL-5 and IL-6 stimulates the

proliferation and differentiation of activated B-cells, andinduces class switch.9,28,29 However, B-cells stimulatedwith IL-5 become plasma cells secreting IgA. IL-5 is also veryimportant in stimulating growth and differentiation ofeosinophils.30e33 The production of cytokines by humanmast cells has not been as extensively studied as in rodents,but several studies suggest that it has a similar pattern. Forexample, human mast cells have been shown to produce IL-4, IL-5, and IL-6.30,31 In addition, mast cells produce severalneutral proteases including tryptase and chymase thatpotentially damage and activate the bronchial epithelium,and may contribute to airway wall remodeling. Thus, mastcells are key players in host defense, with a role in immunesurveillance, phagocytosis, and immune activation.

Functions of mast cells in physiological andpathological states

The biological function of mast cell neutral proteasesremains to be fully clarified. In serum, elevated levels oftryptase are detected in systemic mast cell disorders, suchas anaphylaxis and mastocytosis. Ongoing mast cell activa-tion in asthma appears to be a characteristic of the chronicinflammatory nature of the disease. Activation is detectedby elevated levels of tryptase and PGD2 in bronchoalveolarlavage (BAL) and higher spontaneous release of histamine bymast cells obtained from the BAL of asthmatics than thoseobtained from non asthmatics.21 Ultrastructural analysis ofmast cells in lung tissue also shows that asthmatics havemore degranulation than atopic nonasthmatics.34 Thenumber of the cells increases at sites of inflammation. Toreach these areas, mast cell progenitors must migrate fromthe blood into tissue sites. A crucial step in this process is theadherence of cells to the endothelium. Cell adherence ismediated by several families of adhesion molecules andadhesion receptors on the surface of mast cells that canmediate binding to other cells and to extracellular matrix(ECM) glycoprotein. Upon stimulation, mast cells releasecytokines, including TNF-a and IL-4 that can modulateadhesion molecules on endothelial cells. Activated endo-thelial cells express the intercellular adhesion molecule(ICAM-1), endothelial-leukocyte adhesion molecule-1(ELAM-1) and vascular cell adhesion molecule (VCAM-1) ontheir cell surface.35,36 Human mast cells express integrins asreceptors for these molecules. Until recently, the effects ofadherence on cell function were believed to result only fromchanges in cell shape and cytoskeletal organization.However, in addition to cell spreading, aggregated adhesionreceptors transduce a variety of intracellular signals thatregulate cell function. These signals include protein tyrosinephosphorylation, phosphoinositide hydrolysis and changes inintracellular pH or calcium concentration and the expressionof several genes. The adhesion properties of the cells regu-late their migration, localization, proliferation and pheno-type. Recently, murine mast cells have been implicated inthe mediation of inflammatory responses at long distances.TNF alpha containing particles released by the cells werefound transported through draining lymphatic system acti-vating cells in the lymph nodes.37

Different mechanisms could contribute to the increase inthe number of mast cells at the sites of tissue injury: mast

Figure 2 Induction and effector mechanisms in type I hypersensitivity.

12 K. Amin

cells or their progenitors could migrate to these sites; orresident mast cell precursors could proliferate. Adhesionreceptors and their ligands also play a role in the localizationand migration of mast cells in normal tissues. ECM proteinsthat are the ligands for adhesion receptors are chemotacticfor mast cells. Adherence of mast cells to fibroblasts, othercells or to ECM proteins can transduce signals that affect cellgrowth and differentiation. The increase in the number ofmast cells,38,39 and the enhanced secretion at sites ofinflammation, can accelerate the elimination of the cause oftissue injury or, paradoxically, may lead to a chronicinflammatory response. Thus, manipulating mast cell adhe-sionmay be an important strategy in controlling the outcomeof allergic and inflammatory responses.

In a previous study by us,1 Mast cell numbers wereincreased in both allergic and non-allergic asthma. Similar toamore recent study,we found17 infiltration ofmast cells in ofthe bronchi of both allergic asthmatics and non-allergicasthmatics40 but the accumulation of mast cells was morepronounced for the allergic asthmatics. We did also find thatmast cells in the bronchial mucosa of the allergic asthmaticsshowed more signs of activation with extracellular deposi-tion of tryptase than mast cells from non-allergic asth-matics.1 Signs of mast cell activation in allergic asthma havebeen indicated by others.40 However, in that study we couldnot observe any differences between allergic and non-allergic asthmatics.40 This could be explained by limitednumber of non-allergic asthmatics included in the study.Specific allergens and their reaction with IgE on the mastcells might provide the mechanisms for activation in the

bronchial mucosa in allergic asthma, especially since allthose patients were sensitized to perennial allergens such asfrom pets.41

Also, other authors found inflammatory cells in bronchialmucosa in subjects with toluene Diisocyanate (TDI) inducedasthma.42 The mast cell, which is one of the inflammatorycells, plays an important role in TDI activation because theactivation of mast cells is associated with TDI-induced earlyand late asthmatic reaction.42,43

Mast cells are increased in number in many fibroticdiseases and may play a crucial role in the development offibrosis.44,45 The percentage of humanmast cells in BAL fluidfrompatientswith sarcoidosis or interstitial fibrosis is greaterthan in BAL fluid from healthy individuals,44,46 and patientswith idiopathic interstitial pulmonary fibrosis show evidenceofmast cell degranulation and elevatedmast cell numbers.47

Concluding in a previous study show that Mast cells arelocated in connective tissue, including the lung, skin, thelinings of the stomach and intestine, and other sites. Theyplay an important role in helping defend these tissues fromdiseases. By releasing chemical such as histamine, mastcells attract other key players of the immune defensesystem to areas of the body where they are needed.

Mast cells and airway remodeling

Tissue remodeling is characteristic feature of asthma andother lung diseases. The mechanisms behind this relation-ship between mast cells and fibrosis/tissue remodeling are

Figure 3 Staining of tissue mast cells with anti-tryptase antibody 1 (AA1) in (A) airway smooth muscle (arrow) of patients withallergic asthma and (B) control subjects, respectively. (Mayer’s hematoxylin). Original magnification: �40. The scale bars 50 mm forA and B are shown in the figures.

The role of mast cells in allergic inflammation 13

unclear. We have shown that mast cells may havea substantial effect on tissue remodeling, especially in theairway, on smooth muscle hypertrophy and on mucus hyper-secretion, by releasing proteases such as tryptase, andgrowth factors.17 These cells also have an effect onepithelial damage, and on basement membrane thickeningin patients with allergic asthma,1,48 allergic rhinitis,49 andmiddle ear infection in allergy patients,50,51 mast cellsrelated to airway smooth muscle hypertrophy, compared tohealthy controls.17 Fig. 3 shows that mast cells specificallyare localized within or close to airway smooth musclebundles in patients with allergic asthma whereas little or nomast cells are found in the airways of healthy controls(Fig. 3). Tryptase and other proteases such as chymase areabundant in mast cell granules. Therefore, mast cells seemto play a crucial role in airway remodeling by releasingtryptase onto smooth muscle and epithelium, and may playa role in skin tissue remodeling by releasing chymase in anIgE-dependent manner in allergic diseases.

These data suggest multiple mechanisms and multiplelevels in different organs in the human body, where mastcells can regulate tissue fibrosis and repair, and provideevidence for the direct involvement of mast cells in fibrosisand human connective tissue remodeling.

Conclusions

Mast cells are fascinating, multifunctional, bone marrow-derived, tissue-dwelling cells. They can be activated todegranulate in minutes, not only by IgE and antigen signalingvia the high-affinity receptor for IgE, but also by a diversegroup of stimuli. These cells can release a wide variety ofimmune mediators, including an expanding list of cytokines,chemokines, and growth factors. Mast cells have been shownto play roles in allergic inflammation and, more recently,they have been shown to modulate coagulation cascades,host defense, and tissue remodeling. The role ofmast cells inasthma and other diseases is being actively studied.

This review suggests that mast cells may not onlycontribute to the chronic airway inflammatory response,airway remodeling and symptomatology, but may also haveacentral roleat the initiationof theallergic immuneresponse,that is, providing signals inducing B cell IgE synthesis and Th2lymphocyte differentiation. Th-targeted therapy would be ofconsiderable interest in controlling allergic asthma. Having

more knowledge and resources about mast cells can lead tofinding cures to diseases caused by the mast cells.

Acknowledgments

I am thankful for Professor Godfried M. Roomans, ProfessorChrister Jonson and Dr Jonas Bystrom for taking the time toread and give valuable suggestions to the review. I thank thefollowing institution for kindly giving us permission to publishresults obtained at their sites. Images of bronchial biopsieswere obtained at the Department of Respiratory Medicineand Allergology and Clinical Chemistry at Uppsala AcademicHospital, Uppsala, Sweden. This study was supported by thefoundation Lily and Ragnar Akerhams foundation and grantsfrom the Swedish Heart and Lung Foundation.

References

1. Amin K, Ludviksdottir D, Janson C, Nettelbladt O, Bjornsson E,Roomans GM, Boman G, Seveus L, Venge P. Inflammation andstructural changes in the airways of patients with atopic andnonatopic asthma. BHR group. Am J Respir Crit Care Med 2000;162:2295e301.

2. Ehrlich P. Beitrage zur theorie und praxis der histologischen.Germany: University of Leipzig; 1878.

3. Borish L, Joseph BZ. Inflammation and the allergic response.Med Clin North Am 1992;76:765e87.

4. Blechman JM, Lev S, Brizzi MF, Leitner O, Pegoraro L, Givol D,Yarden Y. Soluble c-kit proteins and antireceptor monoclonalantibodies confine the binding site of the stem cell factor. JBiol Chem 1993;268:4399e406.

5. Ishizaka T, Mitsui H, Yanagida M, Miura T, Dvorak AM. Devel-opment of human mast cells from their progenitors. Curr OpinImmunol 1993;5:937e43.

6. Mitsui H, Furitsu T, Dvorak AM, Irani AM, Schwartz LB,Inagaki N, Takei M, Ishizaka K, Zsebo KM, Gillis S, et al.Development of human mast cells from umbilical cord bloodcells by recombinant human and murine c-kit ligand. Proc NatlAcad Sci USA 1993;90:735e9.

7. Thompson-Snipes L, Dhar V, Bond MW, Mosmann TR, Moore KW,Rennick DM. Interleukin 10: a novel stimulatory factor for mastcells and their progenitors. J Exp Med 1991;173:507e10.

8. Okayama Y, Kawakami T. Development, migration, and survivalof mast cells. Immunol Res 2006;34:97e115.

9. Nakanishi K. Basophils are potent antigen-presenting cells thatselectively induce th2 cells. Eur J Immunol 2010;40:1836e42.

10. Enerback L. Mast cells in rat gastrointestinal mucosa. 2. Dye-binding and metachromatic properties. Acta Pathol MicrobiolScand 1966;66:303e12.

14 K. Amin

11. Oskeritzian CA, Zhao W, Min HK, Xia HZ, Pozez A, Kiev J,Schwartz LB. Surface cd88 functionally distinguishes the MCTCfrom the mct type of human lung mast cell. J Allergy ClinImmunol 2005;115:1162e8.

12. Schwartz LB, Irani AM, Roller K, Castells MC, Schechter NM.Quantitation of histamine, tryptase, and chymase in dispersedhuman T and TC mast cells. J Immunol 1987;138:2611e5.

13. Church MK, Levi-Schaffer F. The human mast cell. J AllergyClin Immunol 1997;99:155e60.

14. Naclerio RM. Pathophysiology of perennial allergic rhinitis.Allergy 1997;52:7e13.

15. Klein LM, Lavker RM, Matis WL, Murphy GF. Degranulation ofhuman mast cells induces an endothelial antigen central toleukocyte adhesion. Proc Natl Acad Sci USA 1989;86:8972e6.

16. Metcalfe DD, Baram D, Mekori YA. Mast cells. Physiol Rev 1997;77:1033e79.

17. Amin K, Janson C, Boman G, Venge P. The extracellulardeposition of mast cell products is increased in hypertrophicairways smooth muscles in allergic asthma but not in nonal-lergic asthma. Allergy 2005;60:1241e7.

18. Ghaffar A. Hypersensitivity reactions. Microbiology andimmunology. Immunology: USC School of Medicine; 2006.

19. Ohkawara Y, Yamauchi K, Tanno Y, Tamura G, Ohtani H,Nagura H, Ohkuda K, Takishima T. Human lung mast cells andpulmonary macrophages produce tumor necrosis factor-alphain sensitized lung tissue after IgE receptor triggering. Am JRespir Cell Mol Biol 1992;7:385e92.

20. Shumilina E, Lam RS, Wolbing F, Matzner N, Zemtsova IM,Sobiesiak M, Mahmud H, Sausbier U, Biedermann T, Ruth P,et al. Blunted IgE-mediated activation of mast cells in micelacking the Ca2þ-activated Kþ channel KCa3.1. J Immunol2008;180:8040e7.

21. Holgate ST In: Middleton Jr E, et al., editors. Allergy, partic-ipates and practice. St-Louis: Mosby; 1993. p. 267.

22. Schwartz L, et al In: Middleton Jr E, et al., editors. Allergy,participates and practice. St-Louis: Mosby; 1993.

23. Stevens RL, Austen KF. Recent advances in the cellular andmolecular biology of mast cells. Immunol Today 1989;10:381e6.

24. Abraham SN, Malaviya R. Mast cells in infection and immunity.Infect Immun 1997;65:3501e8.

25. Arock M, Ross E, Lai-Kuen R, Averlant G, Gao Z, Abraham SN.Phagocytic and tumor necrosis factor alpha response of humanmast cells following exposure to gram-negative and gram-positive bacteria. Infect Immun 1998;66:6030e4.

26. Holgate ST. The epithelium takes centre stage in asthma andatopic dermatitis. Trends Immunol 2007;28:248e51.

27. Kraneveld AD, James DE, de Vries A, Nijkamp FP. Excitatorynon-adrenergic-non-cholinergic neuropeptides: key players inasthma. Eur J Pharmacol 2000;405:113e29.

28. Bradding P. Human mast cell cytokines. Clin Exp Allergy 1996;26:13e9.

29. Bradding P. The role of the mast cell in asthma: a reassess-ment. Curr Opin Allergy Clin Immunol 2003;3:45e50.

30. Bradding P, Feather IH, Wilson S, Bardin PG, Heusser CH,Holgate ST, Howarth PH. Immunolocalization of cytokines inthe nasal mucosa of normal and perennial rhinitic subjects.The mast cell as a source of IL-4, IL-5, and IL-6 in humanallergic mucosal inflammation. J Immunol 1993;151:3853e65.

31. Bradding P, Roberts JA, Britten KM, Montefort S, Djukanovic R,Mueller R, Heusser CH, Howarth PH, Holgate ST. Interleukin-4,-5, and -6 and tumor necrosis factor-alpha in normal and asth-matic airways: evidence for the human mast cell as a source ofthese cytokines. Am J Respir Cell Mol Biol 1994;10:471e80.

32. Butch AW, Chung GH, Hoffmann JW, Nahm MH. Cytokineexpression by germinal center cells. J Immunol 1993;150:39e47.

33. Del Prete G. Human th1 and th2 lymphocytes: their role in thepathophysiology of atopy. Allergy 1992;47:450e5.

34. Djukanovic R, Lai CK, Wilson JW, Britten KM, Wilson SJ,Roche WR, Howarth PH, Holgate ST. Bronchial mucosal mani-festations of atopy: a comparison of markers of inflammationbetween atopic asthmatics, atopic nonasthmatics and healthycontrols. Eur Respir J 1992;5:538e44.

35. Bacon AS, McGill JI, Anderson DF, Baddeley S, Lightman SL,Holgate ST. Adhesion molecules and relationship to leukocytelevels in allergic eye disease. Invest Ophthalmol Vis Sci 1998;39:322e30.

36. Gudbjornsson B, Hallgren R, Nettelbladt O, Gustafsson R,Mattsson A, af Geijerstam E, Totterman TH. Phenotypic andfunctional activation of alveolar macrophages, T lymphocytesand NK cells in patients with systemic sclerosis and primarysjogren’s syndrome. Ann Rheum Dis 1994;53:574e9.

37. Kunder CA, St John AL, Li G, Leong KW, Berwin B, Staats HF,Abraham SN. Mast cell-derived particles deliver peripheralsignals to remote lymph nodes. J Exp Med 2009;206:2455e67.

38. Amin K, Janson C, Harvima I, Venge P, Nilsson G. CC chemokinereceptors CCR1 and CCR4 are expressed on airway mast cells inallergic asthma. J Allergy Clin Immunol 2005;116:1383e6.

39. Brightling CE, Ammit AJ, Kaur D, Black JL, Wardlaw AJ,Hughes JM, Bradding P. The CXCL10/CXCR3 axis mediateshuman lung mast cell migration to asthmatic airway smoothmuscle. Am J Respir Crit Care Med 2005;171:1103e8.

40. Brightling CE, Bradding P, Symon FA, Holgate ST, Wardlaw AJ,Pavord ID. Mast-cell infiltration of airway smooth muscle inasthma. N Engl J Med 2002;346:1699e705.

41. Humbert M, Grant JA, Taborda-Barata L, Durham SR, Pfister R,Menz G, Barkans J, Ying S, Kay AB. High-affinity IgE receptor(FcepsilonRI)-bearing cells in bronchial biopsies from atopicand nonatopic asthma. Am J Respir Crit Care Med 1996;153:1931e7.

42. Park HS, Hwang SC, Nahm DH, Yim HE. Immunohistochemicalcharacterization of the cellular infiltrate in airway mucosa oftoluene diisocyanate (TDI)-induced asthma: comparison withallergic asthma. J Korean Med Sci 1998;13:21e6.

43. Di Stefano A, Saetta M, Maestrelli P, Milani G, Pivirotto F,Mapp CE, Fabbri LM. Mast cells in the airway mucosa and rapiddevelopment of occupational asthma induced by toluene dii-socyanate. Am Rev Respir Dis 1993;147:1005e9.

44. Holdsworth SR, Summers SA. Role of mast cells in progressiverenal diseases. J Am Soc Nephrol 2008;19:2254e61.

45. Levi-Schaffer F, Rubinchik E. Mast cell role in fibrotic diseases.Isr J Med Sci 1995;31:450e3.

46. Chlap Z, Jedynak U, Sladek K. Mast cell: it’s significance inbronchoalveolar lavage fluid cytologic diagnosis of bronchialasthma and interstitial lung disease. Pneumonol Alergol Pol1998;66:321e9.

47. Hunt LW, Colby TV, Weiler DA, Sur S, Butterfield JH. Immuno-fluorescent staining for mast cells in idiopathic pulmonaryfibrosis: quantification and evidence for extracellular releaseof mast cell tryptase. Mayo Clin Proc 1992;67:941e8.

48. Oh CK. Mast cell mediators in airway remodeling. ChemImmunol Allergy 2005;87:85e100.

49. Amin K, Rinne J, Haahtela T, Simola M, Peterson CG,Roomans GM, Malmberg H, Venge P, Seveus L. Inflammatorycell and epithelial characteristics of perennial allergic andnonallergic rhinitis with a symptom history of 1 to 3 years’duration. J Allergy Clin Immunol 2001;107:249e57.

50. Hurst DS, Amin K, Seveus L, Venge P. Mast cells and tryptase inthe middle ear of children with otitis media with effusion. Int JPediatr Otorhinolaryngol 1999;49(Suppl. 1):S315e9.

51. Hurst DS, Amin K, Seveus L, Venge P. Evidence of mast cellactivity in the middle ears of children with otitis media witheffusion. Laryngoscope 1999;109:471e7.

Glucocorticoids: mechanisms ofaction and anti-inflammatorypotential in asthma

V. H. J. van der Velden

Department of Immunology, Erasmus University,PO Box 1738, 3000 DR Rotterdam, The Netherlands

Te l: (+31) 10 408 7192Fax : (+31) 10 436 7601Email: [email protected]

GLUCOCORTICOIDS are potent inh ibitors of in flam m atoryprocesse s and are w ide ly used in th e treatm ent ofas thm a. Th e an ti-inflam m atory effects are m ediatedeither by direct binding of the glucocorticoid/gluco-corticoid receptor com plex to glucocorticoid respon-sive e lem ents in th e prom oter region of genes , or byan in te raction of th is com plex w ith other tran scrip-tion factors , in particular activating prote in-1 ornuclear factor -k B. Glucocorticoids in hibit m anyin flam m ation-associa ted m olecules such as cyto-kines , chem okines , arachidon ic acid m etabolites , andadhesion m olecules . In contras t, an ti-inflam m atorym ediators often are up-regulated by glucocorticoids .In vivo studies have shown that tr eatm ent of asth -m atic patients w ith in haled glucocorticoids in h ibitsthe bronchial in flam m ation and s im ultaneouslyim proves th eir lung function . In th is r evie w , ourcurren t kn owledge of th e m echan ism of action ofglucocor ticoids and their anti-in flam m atory poten tialin as thm a is described. Since bronch ial epithe lialcells m ay be im portan t targe ts for glucocorticoidtherapy in asth m a, the effects of glucocor ticoids onepithe lial ex pre ssed inflam m atory genes w ill beem phasized.

Key words : Glucocorticoids, Asthma, Bronchial epithe lialce lls, Inflammation

Introduction

Glucocorticoids are hormones synthe sized in theadrenal cortex and sec rete d into the blood, w here theleve ls of glucocorticoids fluctuate in a circadianmode . In humans, the naturally occurring gluco-corticoid is hydrocortisone (cortisol), w hich is syn-thesized from its precursor cortisone .

The be neficial e ffec ts of glucocorticoids in as th-matic patients w ere first de scribed in 1950.1 Sincethen on, many studie s have focused on the ther-apeutic potential of glucocorticoids. Several syntheticglucocorticoids, much more potent than cortisol andw ithout the unw anted mine ralocorticoid side e ffec ts,have been deve loped. Now adays, glucocorticoids arepow erful agents in the tre atment of inflammatorydiseases and are by far the most effective anti-inflammatory drugs used in the treatment ofasthma.

Mechanism of ActionAlthough glucocorticoids have been know n for a longperiod of time , the ir prec ise mechanism of ac tion isstill not complete ly unde rstood. However, recentstudies have inc reased our understanding of theircomplex mechanisms of action.

Glucocorticoid receptor

To ex ert the ir effe cts, glucocorticoids ne ed to bind toa spec ific cytoplasmic glucocorticoid receptor (GR).Almost all ce lls of the body ex pre ss the GR, but thenumber of re ceptors may vary be tw een differe nt ce lltype s.2 Cloning of the GR has re vealed that the GRcons is ts of approx imately 800 amino ac id residues,and that ce rtain areas of the molecule show homol-ogy w ith othe r steroid re ceptors, receptors forthyroid hormones, and receptors for re tinoic ac id.3 –7

All members of the nuclear hormone receptor familyshare a charac teristic three-domain structure , firstdesc ribed for the human GR. The C-terminal domainis equal in size in all nuclear receptors studied (about250 amino ac ids) and its main function is to bind thesteroid.8 It also contains the binding site s for the heatshock prote ins (hsp) 90.9,1 0 Removal of the steroid-binding domain results in a constitutively ac tive GRmolecule, indicating that this part of the moleculeacts as a repressor of the transcription-activationfunction. The most conserved central domain isinvolved in direct binding of the re ceptor to DNA. Itcontains tw o distinct loops of prote in, each bound attheir base via four cyste ine re sidue s to a single zincion, the so-called zinc fingers.1 1 These zinc cluste rsare involved in binding of the GR to the major grooveof the DNA double he lix and play a role in dimeriza-

0962-9351/98/040229-09 $9.00 © 1998 Carfax Publishing Ltd 229

Review

Mediators of Inflammation, 7, 229–237 (1998)

tion of tw o GR molecule s.1 2,1 3 In addition, the centralDNA-binding domain has a transcription-activationfunction.4 ,14 The ste roid-binding and DNA-bindingdomains are separated by the ‘hinge-region’, w hichcontains sequence s that are important for nucleartranslocation and dimerization.9,10 The N-te rminaldomain is ex treme ly variable in size (24–600 aminoacids). Its pre cise role is still uncertain, but it isrequired in transc riptional activation.15

Tw o different forms of the human GR have be endesc ribe d.3 ,1 6 These tw o highly homologous isoforms,te rmed GRa and GRb , are gene rated by alternativesplic ing of the human GR pre -mRNA. The GRbisoform diffe rs from the GRa isoform only in itsC-terminal domain, in w hich the last 50 amino ac idsof the latte r are replaced by a unique 15 amino ac idsequence . How ever, this re placement has dramaticfunctional consequence s, since the GRb isoform isunable to bind glucocorticoids and to transduceligand-dependent transac tivation. How eve r, the phys-iological significance of the GRb isoform remainsquestionable, s ince some recent studies indicate thatthis form is not conse rved among spec ie s and nodominant negative inhibition of GRa activity could befound.17 ,1 8 Nevertheless , abundant ex pre ssion of GRbprote in can be found in the epithe lial cells lining thete rminal bronchioli of the lung.19

The ex press ion of the GR may be re gulated bynumerous fac tors e ither at the transc riptional, transla-tional or post-translational leve l.2 0,21 Glucocorticoidshave be en show n to dow n-regulate the ex pression ofthe GR, both in v itro and in vivo .22 ,23 In contrast,inflammatory mediators like interleukin (IL)-1 b , IL-4,tumour ne crosis fac tor (TNF)-a , lipopolysac charide(LPS) and interferon (IFN)-g have been show n toinc re ase glucocorticoid binding in vitro .24 – 28 How -eve r, the inc rease in GR numbers may be ac com-panied by a reduced affinity for glucocorticoids.24 ,2 8

Analysis of GR localization in normal and asthmaticlung has not revealed differenc es in the leve l or sitesof GR ex pre ssion.2 9

Regulation of gene transcription

In the absence of glucocorticoids, the GR is presentin the cytoplasm of the cell as a he te ro-oligomercons is ting of the GR itse lf, tw o molecule s of hsp90, one molecule hsp 70, and one molecule of hsp56 (w hich probably does not interact w ith the GRitself, but interacts w ith hsp 90).30 – 3 4 Glucocor-ticoids enter the cytoplasm of the ce ll by passivediffusion through the cell membrane. In the cyto-plasm they bind to the GR complex , w hich subse-quently undergoes conformational change s, re sult-ing in the dissoc iation of the hsp 90 and hsp 56molecules . Upon this activation, the glucocorticoid-GR complex passe s the nuc lear membrane, entersthe nuc leus, and the hsp 70 molecule is dissoc iated.

Furthe rmore , in the nucleus liganded GR form hom-odimers (Fig. 1).

Within the nucleus, the GR homodimers mayregulate gene transc ription in seve ral ways: (1) viabinding of the glucocorticoid-GR complex to spec ificDNA sequences, the reby directly ac tivating or re pre s-sing gene s; (2) via inte raction w ith other transcriptionfactors; and (3) via modulating the stability of specificmRNA molecule s.3 5 – 3 9

Bin ding to DNA s eque nce sSeve ral steroid-re sponsive genes contain glucocor-ticoid re sponsive e lements (GRE) in their promoterregion.3 5,4 0 Binding of GR homodimers to GRE mayeither re sult in transcriptional activation of the gene(via a po s itive GRE) or repression of the gene (via anega tive GRE) (Fig. 1). The consensus sequence for(positive) GRE is the palindromic 15-base-pairsequence GGTACAnnnTGTTCT, w hereas the negativeGRE has a more variable sequence.36 The rate oftransc riptional regulation of steroid-re sponsive genesis dependent on both the numbers of GRE, the affinityof the glucocorticoid-GR complex to the GRE, and theposition of the GRE re lative to the transcriptional startsite . Binding of the complex to GRE may re sult inconformationa l change s in the DNA and ex posure ofpreviously masked areas, re sulting in increased bind-ing of othe r transcription factors.4 1– 44

In te ra c tio n w ith o th e r tra n s criptio n fa c to r sMany ste roid-responsive gene s do not have GRE intheir promoter region. How ever, binding site s forothe r transcription factors, including nuclear factor(NF)-k B, ac tivating prote in (AP)-1, and cAMP-respon-sive e lement binding protein (CREB), often can befound.45

V. H. J. van de r Ve lden

230 Mediators of Inflammation · Vol 7 · 1998

FIG. 1. Schematic representation of the cellular events afteradministration of glucocorticoids (adapted from Ref. 39).

AP-1, w hich is a dimer of tw o proto-oncogene s(members of the c-jun and c-fos family),46 ,47 isinvolved in the regulation of several genes, includingadhesion molecules and cytokines (review ed in Ref.47). Direct protein–prote in inte raction betw een AP-1and the glucocorticoid-GR complex results in rec ipro-cal repre ssion of one another’s transc riptional activa-tion by pre venting binding of the AP-1 and gluco-corticoid-GR complex to AP-1 sites and GRE,respective ly (Fig. 1).3 7,4 8,4 9

Comparable to AP-1, NF-k B (a hete rodimer of p50and p65 subunits5 0,51 ) re gulate s the transcription ofseveral gene s involved in inflammatory reac-tions.5 0,52,5 3 In unstimulated ce lls , NF-k B is re tained inthe cytoplasm of the cells through the interaction w iththe inhibitor s Ik Ba and Ik Bb .5 4 – 56 Upon cell stimula-tion, for ex ample by IL-1 b or TNF-a , I k B are rapidlyphosphorylated, ubiquitinated, and consequently pro-te olysed.53 ,57 The liberated NF-k B dimers translocateto the nucleus w here they can ac tivate target genes.Glucocorticoids may inhibit NF-k B-stimulated gene sby a direct interaction betw een the glucocorticoid-GRcomplex and the p65 subunit of NF-k B, re sulting intransrepre ssion (Fig. 1).5 1,5 5,5 8,59 Furthe rmore , gluco-corticoids may indirec tly antagonize NF-k B mediate dtransc ription by up-regulating the synthesis of theinhibitory protein Ik Ba , w hich traps NF-k B in inac tivecytoplasmic complex e s.39 ,54 ,5 5 A large number ofimmunoregulatory genes, w hose ex pre ssion isinduced by a variety of pro-inflammatory mediators,contain NF-k B sites in their promote rs/re gulatoryregions. Therefore , it is no w onder that glucocorticoidshave been found to prevent the ex pre ssion of the segenes, including those coding for IL-1 b , IL-2, IL-6, IL-8,monocyte chemoattrac tant prote in (MCP)-1, RANTES(Regulated upon Activation, Normal T ce ll Ex pre ssed,and presumably Sec rete d), granulocyte macrophagecolony-stimulating factor (GM-CSF), the IL-2 receptor,inte rcellular adhe sion molecule (ICAM)-1, andE-se lectin (review ed in Ref. 45). Probably, interactionsbetw een glucocorticoids and NF-k B or AP-1 w illex plain most of the anti-inflammatory and immuno-suppressive activities of glucocorticoids.

An inte rac tion be tw een CREB and the gluco-corticoid-GR complex has also been suggeste d.6 0,6 1

b -agonists, w hich are used as bronchodilators in thetre atment of asthma, inc rease cAMP formation andsubsequently ac tivate CREB. Therefore , s imultaneoustre atment of as thmatic patients w ith glucocorticoidsand b -agonists may result in reduced re sponsivene ssof the airw ays for steroids.6 1– 63

Mo dula tio n o f m RNA s ta bility.A third mechanism by w hich glucocorticoids mayregulate the synthesis of prote ins is via enhancedtransc ription of spec ific ribonuclease s w hich are ableto degrade mRNA containing constitutive AU-richsequence s in the untranslated 3 9 -re gion.64 Such gluco-

corticoid-mediated modulation of post-translationalevents (resulting in decreased mRNA stability andreduced half-life time) has be en observed for IL-1 b , IL-6 and GM-CSF.6 5,6 6

Glucocorticoid Regulated GenesGlucocorticoids are able to modulate the transcrip-tion of a varie ty of gene s, including cytokines andchemokines, re ceptors, enzymes, adhesion mole-cules , and inhibitory prote ins (Table 1). Since epithe-lial c ells may be one of the most important targets forglucocorticoid the rapy in asthma, the effe cts ofglucocorticoids on epithe lial ex pressed inflammatorygenes w ill be emphas ized in this re view.

Cytokines and chemokines

Glucocorticoids inhibit the transcription of mostcytokines and chemokines that are re levant in asthma,including IL-1 b , TNF-a , GM-CSF, IL-3, IL-4, IL-5, IL-6, IL-8, IL-11, IL-12, IL-13, RANTES, eotax in, and mac ro-phage inhibitory prote in (MIP)-1 a .45 ,66 In general,reduced synthe sis of these mediators may re sult in adecre ased re cruitment and ac tivation of leukocytes,also indire ctly due to e ffec ts on adhe sion molecule sand cell survival. Since many cytokine gene pro-mote rs do not contain a ne gative GRE, the effects ofglucocorticoids on cytokine and chemokine produc-tion are probably mediated via an e ffec t on a critic altransc ription factor (e spec ially NF-k B and AP-1).6 7

Cross-signalling betw een NF-k B and AP-1 w ith gluco-corticoid/GR complex have indeed be en demon-strated in bronchial ep ithe lial cells.6 7

Glu co co rtico id a ctio n in a s th m a

Mediators of Inflammation · Vol 7 · 1998 231

Table 1. Influence of glucocorticoids on the synthesis ofproteins with inflammatory effects by bronchial epithelialcells

Protein Glucocorticoideffect

CytokinesIL-1b , IL-6, IL-11, TNF- a , GM-CSF ¯IL-10, LIF ?G-CSF =

ChemokinesMCP-1, eotaxin, IL-8, RANTES, MIP-1a ¯

ReceptorsNK, GR ¯IL-1R II, IL-6R, b 2-adrenergic receptor

EnzymesiNOS, COX-2, cPLA2 ¯NEP

Adhesion moleculesICAM-1 ¯

Inhibitory proteinsLc-1 =/IL-1RA type I, SPLI

Bronchial epithelial cells are capable of producing avarie ty of cytokine s and chemokine s that may contrib-ute to the initiation and pe rpetuation of airw ayinflammation. Several studies have show n that cyto-kine-induced ex pression of eotax in, IL-6, IL-8, GM-CSF, and RANTES can be diminished by gluco-corticoids in v itro .68 –76 In contrast, glucocorticoidsdid not modulate the sec re tion of G-CSF by humanbronchial epithelial cells.7 6 In v ivo studie s haveshow n that treatment w ith inhaled steroids decre ase sboth the ex press ion of GM-CSF,77 IL-1 b ,78 IL-8,79 andRANTES80 by the bronchial ep ithe lium, toge the r w iththe number of ac tivated eosinophils in theepithe lium.

Receptors

Glucocorticoids may modulate the ex press ion ofseveral re ceptors. The ex pre ssion of the neurokinin(NK)1 re ceptor, w hich mediate s many effects ofsubstance P (SP) in the airw ays and is be lieved to beup-regulated in asthma,8 1 is dow n-re gulated by gluco-corticoids.8 2 Since the NK1 re ceptor gene promoterregion has no GRE but has an AP-1 re sponse e lement,this e ffec t probably w ill be mediated via an inte r-action of the glucocorticoid-GR complex w ith AP-1.

In contrast to NK1 re ceptors, ex press ion of theb 2-adrene rgic receptor is increased by glucocor-tic oids.8 3 Since the human b 2-adrene rgic receptorgene contains three potential GRE, this effect ofglucocorticoids probably is a dire ct one .83 Up-regula-tion of b 2-adrene rgic re ceptors by glucocorticoidsmay be re levant in as thma as it may prevent dow n-regulation in re sponse to prolonged tre atment w ithb 2-agonists .84

The IL-1 receptor type II, w hich func tions as adecoy re ceptor,8 5 may also be up-regulated by gluco-corticoids, there by re ducing the functional ac tivity ofIL-1 agonists.8 6,8 7 Soluble TNF-re ceptor type I (p55)re lease by human bronchial epithelial c ells, bothconstitutive as w ell as IL-1 b -induced, has been show nto be re duced by glucocorticoids.8 8 In contrast,glucocorticoids up-re gulate the ex pre ssion of IL-6receptors in rat hepatoma and human epithelialce lls .89 ,90 Thus far little is know n about this proce ss inhuman bronchial epithe lial ce lls, w hich constitutivelyex pre ss these re ceptors.91

Glucocorticoids also modulate the ex pre ssion oftheir ow n receptor. In a re cent study it w as show nthat ex press ion of the a -form (but not the b -form) ofthe GR in human bronchial ep ithe lial ce lls w as dow n-regulated in healthy subjects after 4 weeks of budeso-nide inhalation.2 3

Enzymes

Glucocorticoids inhibit the synthe sis of severalinflammatory mediators implicated in the pathogene-

sis of asthma through an inhibitory effect on enzymeinduction. The synthe sis of inducible nitric ox idesynthase (iNOS) by human airw ay epithelial cells isinhibited by glucocorticoids, both in vitro and inv ivo .92 – 9 4 This effe ct seems to be mediated viainactivation of NF-k B.9 5,9 6 Since nitric ox ide (NO)may contribute to skew ing of Th lymphocytestow ards a Th2 phenotype , there by promoting IgEproduction and eosinophil rec ruitment, inhibition ofiNOS may be of importance in anti-inflammatorytherapy in as thma.9 7

Glucocorticoids also inhibit the gene transcriptionof a cytosolic form of phospholipase A2 induced bycytokine s98 and inhibit the gene ex pre ssion ofcycloox ygenase-2, re sulting in reduced formation ofprostaglandins and thrombox anes.99

In contrast to the enzymes mentioned above,glucocorticoids have be en show n to increase theex pre ssion of neutral endopeptidase (NEP),10 0 –10 2

thereby pote ntially limiting neurogenic inflammatoryresponses.1 03 In ac cordance w ith the se results, it w asfound that the ex press ion of NEP on bronchialepithe lial c ells w as higher in as thmatic s treatedw ith steroids compared w ith nonste roid-treatedasthmatics.1 04

Endothelins

Endothelins are a family of highly homologous21-amino ac id peptide s, charac terized by tw o intra-chain disulphide chains, a hairpin loop cons isting ofpolar amino ac ids, and a hydrophobic C-te rminalchain.10 5 Human bronchial epithelial c ells have be enshow n to produce ET-1,1 06 –1 08 w hich promotes theprolife ration of smooth muscle ce lls , is a pote ntconstrictor of both vascular and non-vascular smoothmuscle ce lls , inc reases the se cretion of mucus, andmay ac tivate inflammatory ce lls.1 05 ,1 07 ,10 9 ET-1 alsostimulates collagen gene ex pression and through itsinhibitory ac tions on collagenase w ill promote airw ayw all collagen deposition, there by contributing toairw ay w all thickening w hich underlies bronchialhyperresponsiveness.11 0 –11 2 Inc re ased leve ls of ET-1-immunoreac tivity we re detected in airw ay epithe-lium and vascular endothelium of bronchial biopsyspecimens from nonste roid-tre ated asthmatic s com-pared w ith healthy subjects .10 6,1 13 ,11 4 In contrast, noincre ased ET-1 ex pression w as found in the bronchialepithe lium of asthmatic patients treated w ithglucocorticoids.11 5

Adhesion molecules

Adhesion molecules play an important role in therec ruitment of inflammatory ce lls to the inflamma-tory locus. Ex pression of adhesion molecule s onendothelial, epithelial or inflammatory ce lls is often



induced by cytokine s, w hereas glucocorticoidsreduce surface ex pression of adhe sion molecules.This effect may be due either to inhibition of cytokinesynthesis or to a direc t effect of glucocorticoids onadhesion molecule gene transcription. It has be enshow n that the ex pression of ICAM-1, endothelialleukocyte adhesion molecule (ELAM)-1, and E-selec tinis dow n-re gulated by ste roids.11 6 Basal and cytokine-stimulated ICAM-1 ex pre ssion on human bronchialepithe lial c ell lines is inhibited by glucocor-tic oids.1 17 ,11 8 Howeve r, inhaled glucocorticoids didnot modulate ICAM-1 ex pre ssion by bronchial epithe-lial ce lls from asthmatic s in vivo .11 9

Eosinophil adhe sion to cytokine-stimulated bron-chial ep ithe lial c ells w as show n to be inhibited by thesynthetic glucocorticoid dex amethasone .120 Althoughcytokine-activated epithelial c ells show ed increasedex pre ssion of ICAM-1, this molecule did not seem tobe involved in the dec reased adhe sion of eosinophilsobserved in the presenc e of dex amethasone .12 0

Inhibitory proteins

The anti-inflammatory e ffec ts of glucocorticoids maybe mediated by increasing the production of inhibi-tory prote ins, such as lipocortins. Lipocortins aremembers of a supe rfamily of proteins characterizedby the ir ability to bind calcium and anionic phospho-lipids, now know n as the ‘annex ins’.12 1,12 2 In severalce ll type s, but not all, glucocorticoids are inducers oflipocortins, w hich have an inhibitory effec t on theactivity of phospholipase A2 .123 ,1 24 As a re sult, thesynthesis of lipid mediators, including prostaglandinsand e icosanoids, w ill be re duced. How ever, in humanbronchial epithelial c ells glucocorticoids do not se emto up-regulate the ex pre ssion of lipocortins.12 5 Fur-thermore , no significant difference w as foundbetw een lipocortin-1 concentration in BAL fluid fromasthmatic patients re ce iving inhaled glucocorticoidtherapy and those w ho were not treated w ithglucocorticoids.12 6

Recently, glucocorticoids have also been show n toinc re ase the ex pre ssion of intracellular IL-1 receptorantagonist type I by human bronchial epithelial c ellsin vitro .1 27 Increased production of this mediatormay inhibit the effe cts of IL-1 agonists , the rebyreducing inflammation. Howeve r, glucocorticoidtreatment of asthmatic patients did not affe ct theex pre ssion of IL-1 receptor antagonist by the bron-chial epithe lium.78

To provide prote ction against potentially injuriousagents, airw ay epithe lial cells se cre te a number ofmediators, including antiprote ase s. Secretory leuko-cyte protease inhibitor (SLPI) is the pre dominantantiprotease in the airw ays. Its ex pre ssion has be enshow n to be inc reased in airw ay epithelial ce lls afterstimulation w ith glucocorticoids.12 8

Cellular and Clinical Effects ofGlucocorticoids in AsthmaSeve ral s tudie s have determined the effe cts of inhaledglucocorticoids on bronchial inflammation, e ithe r bymeasurements in BAL fluid, sputum, or ex haled air, orby performing bronchial biopsie s. Although differ-ences can be observed betw een differe nt trials , the sestudies have confirmed that glucocorticoid treatmentof asthmatic patients reduces the number and activa-tion of inflammatory ce lls in the airw ays, togetherw ith an improvement of lung func tion. Now adays,the potent anti-inflammatory ac tions of glucocor-ticoids are thought to underlie the clinical e ffic acy oforal glucocorticoids.1 29

Effects of glucocorticoids on immunopathology

Inhaled glucocorticoids decre ase the number andactivation status of most inflammatory cells in thebronchus, including mast cells, dendritic ce lls , eosino-phils , and T lymphocytes . Changes in cellular infiltra-tion are ac companied by modulated ex press ion ofseveral cytokines. Inhaled glucocorticoids have be enshow n to decrease mRNA ex pre ssion of GM-CSF, IL-13, IL-4, and IL-5, w hereas mRNA leve ls of IL-12 andIFN-g inc reased, sugge sting a shift from a Th2-tow ards a more Th1-like environment.77 ,13 0,1 31

Glucocorticoid tre atment is as sociated w ith areduction in m a st c e ll numbers in the bron-chus7 9,1 29 ,1 32 –13 5 and w ith reduced mast ce ll asso-ciated mediators in BAL fluid.13 5,1 36 This may be dueto a re duc tion in IL-3 and ste m ce ll factor production,w hich are necessary for the survival of mast c ells intissue. The (IgE-dependent) re lease of mediators frommast cells does not se em to be affec ted by gluco-corticoid treatment.1 37,13 8

Dendritic cells play an important role in presentingantigens to (naive) T ce lls .139 ,1 40 Inhaled gluco-corticoids have been show n to re duce the numberof dendritic c ells in the human bronchialepithe lium.1 41

Increased numbers of eos inophils are a prominentfeature of asthmatic airw ays.14 2 –148 In v itro studie shave show n that many eosinophil functions, includ-ing adhe renc e and chemotax is, are diminished follow -ing glucocorticoid treatment.1 38 How ever, most datasuggest that eos inophil re sponses to steroids are likelyto be indire ct, s ince eosinophil func tion is markedlyaffec ted by cytokine s e laborated from T lymphocytes(IL-3, IL-4, IL-5, GM-CSF), endothelial c ells (GM-CSF)and epithelial c ells (GM-CSF).1 49 –1 53 In vivo studie sindicate that tre atment w ith inhaled steroids re ducesthe number of eosinophils and eosinophil-re latedmediators in BAL fluid79 ,1 48,15 4 and the numberof (activated) eosinophils in bronchial biop-sie s.79 ,1 29 ,13 2,1 33 ,1 55 Recently, induced sputum hasbeen suggeste d as a useful tool for evaluating the



christinehurd

Highlight

effects of the rapy on airw ay mucosal inflammation.Thus far, most studie s have focused on the pre senc eof eosinophils and eosinophil-re lated mediators . Inaccordance w ith the findings in BAL fluid andbronchial biopsie s, glucocorticoid treatment w asassociated w ith a re duction in sputum eosinophilnumbers , eosinophil cationic protein (ECP), andeosinophil pe rox idase (EPO).156

Glucocorticoids also re duce the number of acti-vated T lymphocyte s in bronchial biopsie s and BALfluid.1 29 ,13 3,1 34 ,1 55,1 5 7 In addition, inhaled corticoste-roids reduced the number of ce lls ex pressing mRNAfor IL-4 or IL-5, and increased the number of cellsex pre ssing mRNA for IFN-g ,1 31 ,13 2 the reby favouringthe deve lopment of Th1 cells.15 8

In addition to the e ffec ts of glucocorticoids onepithe lial ce lls described above, inhaled glucocor-ticoid therapy has be en show n to reduce the shed-ding of epithe lial cells.1 55 ,15 9,16 0 No consistent e ffec tof corticosteroids on the thickness of the basementmembrane has been obse rved.79 ,16 0,161

Besides the suppre ssive e ffec ts on inflammatoryce lls , inhaled glucocorticoids have also show n toinhibit mucus secre tion and mic rovascular leakage (asdete rmined by the dow n-re gulation of plasma pro-te ins in BAL fluid).1 60 ,16 2 –16 6 At present it is not clearw hether this is mediated via a dire ct effe ct ofglucocorticoids on endothe lial or mucous cells, or viaa reduction of inflammatory mediators that inc re asemucus se cretion and vascular leakage.

Effects of glucocorticoids on lung function

Treatment w ith glucocorticoids has be en consiste ntlyshow n not only to re duce the symptoms of as thma,but also bronchial hyperresponsiveness.13 4,16 7–16 9 Incontrast to the rapid inhibitory effe cts of b 2-agonis ts,glucocorticoids given in a s ingle dose are not effe ctivein pre venting early alle rgen-invoked bronchoconstric-tion, but inhibition of the late re sponse has be enclearly demonstrated.1 70 ,17 1 In contrast, chronic treat-ment w ith either oral or inhaled steroids attenuate seven the early bronchoconstric tion to alle rgen,17 1–17 3

an effe ct that probably is mediated via the anti-inflammatory actions of glucocorticoids alre adydesc ribed. Although inhaled glucocorticoids con-sis tently re duce airw ay hyperreac tivity in asthmat-ics ,16 9 even after several months of treatment re spon-siveness fails to re turn to the normal range. This mayreflec t persis tence of structural changes that cannotbe re versed by steroids (such as the thickening of thebasement membrane ), despite of suppress ion of theinflammatory and immunologic al processes.

Concluding RemarksGlucocorticoids are w idely used in the tre atment ofasthma and have anti-inflammatory e ffec ts. The se

effects are mediated either by direc t binding of theglucocorticoid/GR complex to GRE in the promoterregion of respons ive gene s, or by an inte raction of thiscomplex w ith transcription fac tors such as AP-1 andNF-k B. Glucocorticoids inhibit the ex pression of alarge number of inflammation-assoc iated molecules,including cytokines, chemokines, arachidonic ac idmetabolite s, and adhesion molecules. These e ffec tspredominantly are mediated via inhibition of NF-k Bactivity. In contrast, anti-inflammatory mediators,such as NEP and IL-1 receptor antagonist, often areup-regulated by glucocorticoids. The bene fic ial e ffec tsof glucocorticoid the rapy in asthma is demonstratedby in vivo studie s show ing that treatment of as th-matic patients w ith inhaled glucocorticoids inhibitsthe inflammation of the airw ays and s imultaneouslyimprove s the ir lung func tion. The se effe cts may bemediated in part by modulation of epithelial ce llfunctions, since many studies, both in v itro and inv ivo , have show n that glucocorticoids are able tomodulate the inflammatory func tions of bronchialepithe lial ce lls. Further studies on the mechanism ofaction of glucocorticoids w ill e ventually le ad to thedeve lopment of drugs w hich spec ifically inhibit thetransc ription of inflammatory genes w ithout havingne gative side e ffec ts, and w ill contribute to a moreeffic ient tre atment of asthmatic patients .

References1. Carrye r HM, Koelsche GA, Prickman LE, Maytum CK, Lake CF, Williams

HL. Effects of cortisone on bronchial asthma and hay fever occ uring insubjects sens itive to ragw ee d polle n. Pro c Sta ff Me e t, Mayo Clin 1950;25: 482–486.

2. Barne s PJ, Pede rsen S. Effic acy and safe ty of inhaled corticos te roids inasthma. Am Rev Re s pir Dis 1993; 148 : S1 –S26.

3. Hollenbe rg SM, Weinbe rge r C, Ong ES, e t a l. Primary s tructure andex press ion of a func tional human gluc oc or tic oid rece ptor cDNA.Na tu re 1985; 318 : 635–641.

4. Hollenbe rg SM, Giguere V, Segui P, Evans RM. Coloc alization of DNA-binding and transcriptional activation functions in the human gluco-corticoid rece ptor. Ce ll 1987; 49: 39–46.

5. Kumar V, Green S, Staub A, Chambon P. Localisation of the oe s tradiol-binding and putative DNA-binding domains of the human oe strogenre ceptor. EMBO J 1986; 5 : 2231–2236.

6. Rusconi S, Yamamoto KR. Functional dissection of the hormone andDNA binding activ itie s of the glucocorticoid re ceptor. EMBO J 1987; 6 :1309–1315.

7. Evans RM. The ste roid and thyroid hormone re ceptor superfamily.Scie n c e 1988; 240 : 889–895.

8. Evans RM, Hollenberg SM. Cooperative and positional independenttrans-activation domains of the human glucocorticoid receptor. Co ldSprin g Ha rb Sy m p Qu a nt Bio l 1988; 53: 813–818.

9. Picard D, Yamamoto KR. Tw o signals mediate hormone-dependentnuclear localization of the glucocortic oid rec eptor. EMBO J 1987; 6 :3333–3340.

10. Kumar V, Chambon P. The estrogen rec eptor binds tightly to itsre sponsive e lement as a ligand-induce d homodimer. Ce ll 1988; 55 :145–156.

11. Freedman LP, Luisi BF, Korszun ZR, Basavappa R, Sigle r PB, YamamotoKR. The func tion and structure of the me tal coordination site s w ithinthe gluc ocorticoid rec eptor DNA binding domain. Na tu re 1988; 334 :543–546.

12. Freedman LP. Anatomy of the ste roid rece ptor zinc finger re gion.Endo c r Rev 1992; 13: 129–145.

13. Luis i BF, Xu WX, Otw inow ski Z, Fre edman LP, Yamamoto KR, Sigle r PB.Crys tallographic analysis of the interaction of the glucocorticoidre ceptor w ith DNA. Na tu re 1991; 352 : 497–505.

14. Schena M, Fre edman LP, Yamamoto KR. Mutations in the glucocorticoidre ceptor zinc finger re gion that distinguish interdigitated DNA bindingand transc riptional enhance ment activitie s. Ge n e s De v 1989; 3 :1590–1601.



christinehurd

Highlight

15. de Waal RM. The anti-inflammatory activity of glucocortic oids. Mo l Bio lRep 1994; 19 : 81–88.

16. Encio IJ, De te ra-Wadle igh SD. The genomic s tructure of the humanglucoc ortic oid rec eptor. J Bio l Ch em 1991; 266 : 7182–7188.

17. Otto C, Reichardt HM, Schutz G. Absence of glucocorticoid rec eptor-beta in mic e . J Bio l Che m 1997; 272 : 26665 –26668.

18. Hecht K, Carls tedt-Duke J, Stie rna P, Gustafsson J, Bronnegard M,Wikstrom AC. Evidenc e that the be ta-isoform of the human gluco-corticoid rec eptor doe s not ac t as a phys iologically significantre pres sor. J Bio l Che m 1997; 272 : 26659–26664.

19. Oakley RH, We bste r JC, Sar M, Parker CR, Jr, Cidlow ski JA. Ex press ionand subce llular dis tribution of the be ta-isoform of the human gluco-corticoid rece ptor. End o c rin o lo g y 1997; 138 : 5028 –5038.

20. Burns te in KL, Cidlow ski JA. The dow n side of glucocorticoid rec eptorre gulation. Mo l Ce ll En do crino l 1992; 83: C1–C8.

21. Bronnegard M. Steroid re ceptor number. Indiv idual variation anddow nre gulation by tre atment. Am J Re s p ir Crit Ca re Med 1996; 154 :S28–S32.

22. Rosew icz S, McDonald AR, Maddux BA, Goldfine ID, Miesfe ld RL,Logsdon CD. Mechanism of glucocorticoid re ceptor dow n-regulationby glucocortic oids. J Bio l Ch em 1988; 263 : 2581–2584.

23. Korn SH, Woute rs EF, We sse ling G, Arends JW, Thunnissen FB. In vitroand in v ivo modulation of alpha- and be ta-gluc oc orticoid-rece ptormRNA in human bronchial epithe lium. Am J Re s pir Crit Ca re Med1997; 155 : 1117–1122.

24. Kam JC, Sze f le r SJ, Surs W, Sher ER, Leung DY. Combination IL-2 and IL-4re duc es glucoc ortic oid rec eptor-binding affinity and T c e ll response toglucoc ortic oids. J Im m u n o l 1993; 151 : 3460–3466.

25. Rakasz E, Gal A, Biro J, Balas G, Falus A. Modulation of glucocorticos-te roid binding in human lymphoid, monoc ytoid and he patoma ce llline s by inflammatory c ytokines interleukin (IL)-1 beta, IL-6 andtumour ne cros is factor (TNF)-alpha. Sca n d J Im m u n o l 1993; 37 :684–689.

26. Salkow ski CA, Voge l SN. Lipopolysac charide incre ases glucocorticoidre ceptor ex pre ssion in murine macrophages . A possible me chanism forglucoc ortic oid-mediated suppress ion of e ndotox icity. J Im m u no l 1992;149 : 4041–4047.

27. Salkow ski CA, Voge l SN. IFN-gamma mediates incre ased glucocorticoidre ceptor ex press ion in murine macrophages. J Im m u n o l 1992; 148 :2770–2777.

28. Ve rheggen MM, van Hal PT, Adriaansen-Soeting PW, e t a l. Modulation ofglucoc ortic oid rece ptor ex pre ssion in human bronchial epithe lial ce llline s by IL-1 beta, TNF-alpha and LPS. Eu r Re s p ir J 1996; 9 :2036–2043.

29. Adcock IM, Gilbey T, Ge lder CM, Chung KF, Barne s PJ. Glucocorticoidre ceptor loc alization in normal and asthmatic lung. Am J Re s p ir CritCa re Med 1996; 154 : 771–782.

30. Gustafs son JA, Carls tedt-Duke J, Poe llinger L, e t a l. Biochemistry,mole cular biology, and phys iology of the glucocorticoid re ceptor.Endo c r Rev 1987; 8 : 185–234.

31. Mendel DB, Orti E. Isoform compos ition and stoichiome try of theapprox imate ly 90-kDa he at shock prote in assoc iate d w ith gluco-corticoid rece ptors. J Bio l Ch em 1988; 263 : 6695–6702.

32. Smith DF, Stensgard BA, We lch WJ, Toft DO. Assembly of proge ste ronere ceptor w ith heat shock prote ins and rec eptor activation are ATPmediated events. J Bio l Che m 1992; 267 : 1350–1356.

33. Sanchez ER. Hsp56: a nove l heat shock prote in assoc iated w ithuntransformed s te roid re ceptor complex es. J Bio l Che m 1990; 265 :22067–22070.

34. Lebeau MC, Massol N, Herrick J, e t a l. P59, an hsp 90-binding prote in.Cloning and sequenc ing of its cDNA and prepration of a peptide-directed polyc lonal antibody. J Bio l Ch em 1992; 267 : 4281–4284.

35. Truss M, Beato M. Steroid hormone rec eptors: interac tion w ithdeox yribonucle ic acid and transcription fac tors . En do cr Rev 1993; 14 :459–479.

36. Beato M. Gene regulation by ste roid hormone s. Ce ll 1989; 56:335–344.

37. Schule R, Rangarajan P, Kliew er S, e t a l. Functional antagonism be tw eenoncoprote in c-Jun and the gluc oc orticoid rece ptor. Ce ll 1990; 62 :1217–1226.

38. Strahle U, Schmid W, Schutz G. Syne rgistic action of the glucocorticoidre ceptor w ith transcription factors. EMBO J 1988; 7: 3389–3395.

39. Didonato JA, Saatc ioglu F, Karin M. Mole cular mechanisms of immuno-suppre ssion and anti-inf lammatory activitie s by glucocorticoids. Am JRe s pir Crit Ca re Med 1996; 154 : S11–S15.

40. Parke r M. Enhance r e lements activated by ste roid hormones? Na tu re1983; 304 : 687–688.

41. Becke r P, Renkaw itz R, Schutz G. Tissue -spec ific DNaseI hypersens itivesite s in the 5 9 -f lanking se quence s of the tryptophan ox ygenase and thetyrosine aminotrans ferase gene s. EMBO J 1984; 3: 2015–2020.

42. Fritton HP, Igo-Kemenes T, Now ock J, Strech-Jurk U, Theise n M, Sippe lAE. Alternative se ts of DNase I-hyperse nsitive s ite s charac te rize thevarious func tional state s of the chicken lysozyme gene . Na tu re 1984;311 : 163–165.

43. Zaret KS, Yamamoto KR. Revers ible and pers is tent change s in chroma-tin structure accompany activation of a glucoc ortic oid-depe ndentenhance r e lement. Ce ll 1984; 38: 29–38.

44. Pe rlmann T, Wrange O. Spec ific glucocorticoid rece ptor binding to DNAre constituted in a nucle osome. EMBO J 1988; 7: 3073–3079.

45. Barne s PJ. Mole cular me chanisms of ste roid ac tion in asthma. J Alle rg yClin Im m u no l 1996; 97: 159–168.

46. Ransone LJ, Ve rma IM. Nuclear proto-onc oge ne s fos and jun. Annu Re vCe ll Bio l 1990; 6: 539 –557.

47. Ange l P, Karin M. The role of Jun, Fos and the AP-1 c omplex in ce ll-prolife ration and transformation. Bio c him Bio phy s Ac ta 1991; 1072 :129–157.

48. Jonat C, Rahmsdorf HJ, Park KK, e t a l. Antitumor promotion andantiinf lammation: dow n-modulation of AP-1 (Fos/Jun) activ ity byglucoc ortic oid hormone. Ce ll 1990; 62: 1189–1204.

49. Yang-Yen HF, Chambard JC, Sun YL, e t a l. Transcriptional interfe rencebetw een c-Jun and the glucocortic oid re ceptor: mutual inhibition ofDNA binding due to dire ct prote in-prote in interaction. Ce ll 1990; 62:1205–1215.

50. Lenardo MJ, Baltimore D. NF-kappa B: a p le iotropic mediator ofinducible and tis sue-spec ific gene control. [Re view ]. Ce ll 1989; 58 :227–229.

51. Ray A, Pre fontaine KE. Phys ic al assoc iation and functional antagonismbetw een the p65 subunit of transcription factor NF-kappa B and theglucoc ortic oid rec eptor. Pro c Na tl Aca d Sc i USA 1994; 91 :752–756.

52. Grilli M, Chiu JJ, Lenardo MJ. NF-kappa B and Rel: partic ipants in amultiform transcriptional regulatory sys tem. In t Rev Cy to l 1993; 143 :1–62.

53. Sie benlis t U, Franzoso G, Brow n K. Structure , regulation and functionof NF-kappa B. An nu Rev Ce ll Bio l 1994; 10: 405–455.

54. Auphan N, DiDonato JA, Rose tte C, Helmberg A, Karin M. Immunosup-pre ssion by gluc oc orticoids : inhibition of NF-kappa B ac tivity throughinduction of I kappa B synthes is . Sc ie n ce 1995; 270 : 286–290.

55. Scheinman RI, Cogsw ell PC, Lofquis t AK, Baldw in AS Jr. Role oftransc riptional activation of I kappa B alpha in mediation of immuno-suppre ssion by glucocorticoids. Scie n ce 1995; 270 : 283–286.

56. Haskill S, Beg AA, Tompkins SM, e t a l. Characte rization of an immediate-early gene induce d in adherent monoc ytes that encodes I kappa B-likeactivity. Ce ll 1991; 65 : 1281–1289.

57. Finc o TS, Baldw in AS. Me chanistic aspects of NF-kappa B regulation: theemerging role of phosphorylation and prote olys is . Im m un ity 1995; 3 :263–372.

58. Mukaida N, Morita M, Ishikaw a Y, e t a l. Nove l mechanism ofglucoc ortic oid-mediated gene repres sion. Nuclear fac tor-kappa B istarget for gluc ocorticoid-mediated interle ukin 8 gene repress ion. J Bio lChe m 1994; 269 : 13289–13295.

59. Calde nhoven E, Liden J, Wis sink S, e t a l. Negative cros s-talk betw ee nRe lA and the gluc ocorticoid rec eptor: a poss ible mechanism for theantiinf lammatory ac tion of glucocorticoids. Mo l En do crin o l 1995; 9:401–412.

60. Yamamoto KK, Gonzalez GA, Menze l P, Rivier J, Montminy MR.Charac te rization of a bipartite activator domain in transc ription fac torCREB. Ce ll 1990; 60: 611–617.

61. Adcock IM, Stevens DA, Barne s PJ. Interactions of glucoc ortic oids andbeta 2-agonis ts. Eu r Re s p ir J 1996; 9: 160–168.

62. Adcock IM, Pete rs MJ, Brow n CR, Ste vens DA, Barnes PJ. Highconc entrations of beta-adrene rgic agonists inhibit DNA binding ofglucoc ortic oids in human lung in vitro. Bio che m So c Tra n s 1995; 23 :217S.

63. Pe te rs MJ, Adcock IM, Brow n CR, Barnes PJ. Be ta-adrenoceptor agonistsinterfe re w ith glucoc ortic oid re ceptor DNA binding in rat lung. Eu r JPha rm a co l 1995; 289 : 275–281.

64. Pe ppe l K, Vinc i JM, Baglioni C. The AU-rich sequences in the 3 9untranslate d region mediate the increase d turnove r of interfe ron mRNAinduced by glucocorticoids. J Exp Med 1991; 173 : 349–355.

65. Lee SW, Tsou AP, Chan H, e t a l. Glucoc or tic oids se lective ly inhibit thetransc ription of the interleukin 1 beta gene and de cre ase the s tability ofinterleukin 1 be ta mRNA. Pro c Na tl Aca d Sci USA 1988; 85 :1204–1208.

66. Schw iebert LM, Ste llato C, Schle ime r RP. The e pithe lium as a target ofglucoc ortic oid action in the tre atment of asthma. Am J Re s p ir Crit Ca reMed 1996; 154 : S16–S19.

67. LeVan TD, Behr FD, Adkins KK, Miesfe ld RL, Bloom JW. Glucocorticoidre ceptor s ignaling in a bronchial epithe lial ce ll line . Am J Phy s io l 1997;272 : L838–L843.

68. Cox G, Ohtoshi T, Vancheri C, e t a l. Promotion of e os inophil survival byhuman bronchial epithe lial c e lls and its modulation by ste roids. Am JRe s pir Ce ll Mo l Bio l 1991; 4: 525–531.

69. Marini M, Vittori E, Hollemborg J, Mattoli S. Ex press ion of the potentinflammatory cytokines , granulocyte -macrophage-c olony-stimulatingfac tor and interleukin-6 and interle ukin-8, in bronchial epithe lial ce llsof patients w ith asthma. J Alle rg y Clin Im m u no l 1992; 89:1001–1009.

70. Berkman N, Robichaud A, Krishnan VL, e t a l. Ex pres sion of RANTES inhuman airw ay epithe lial ce lls: e ffec t of corticosteroids and inte rleukin-4, –10 and –13. Im m u no lo g y 1996; 87: 599–603.

71. Ste llato C, Beck LA, Gorgone GA, e t a l. Ex pre ssion of the che mokineRANTES by a human bronchial epithe lial ce ll line . Modulation bycytokine s and gluc oc orticoids . J Im m u no l 1995; 155 : 410–418.



72. Kw on OJ, Au BT, Collins PD, e t a l. Inhibition of interle ukin-8 ex pres sionby dex ame thasone in human c ultured airw ay epithe lial ce lls. Im m un o l-o g y 1994; 81: 389–394.

73. Lilly CM, Nakamura H, Kesse lman H, e t a l. Ex press ion of e otax in byhuman lung epithe lial c e lls: induc tion by c ytokines and inhibition byglucoc ortic oids. J Clin Inve s t 1997; 99 : 1767–1773.

74. Bedard M, Mc Clure CD, Schille r NL, Franc oeur C, Cantin A, De nis M.Re le ase of inte rleukin-8, interleukin-6, and c olony-stimulating fac tors byupper airw ay e pithe lial ce lls: implications for cystic fibrosis. Am JRe s pir Ce ll Mo l Bio l 1993; 9: 455–462.

75. Kw on OJ, Jose PJ, Robbins RA, Schall TJ, Williams TJ, Barnes PJ.Glucocortic oid inhibition of RANTES ex pre ssion in human lungepithe lial ce lls. Am J Re s pir Ce ll Mo l Bio l 1995; 12: 488–496.

76. Levine SJ, Larivee P, Logun C, Angus CW, She lhame r JH. Corticos te -roids diffe re ntially re gulate sec retion of IL-6, IL-8, and G-CSF by ahuman bronchial epithe lial ce ll line . Am J Phy s io l 1993; 265 :L360–L368.

77. Sousa AR, Poston RN, Lane SJ, Nakhosteen JA, Lee TH. Detec tion of GM-CSF in asthmatic bronchial e pithe lium and decrease by inhale dcorticos te roids. Am Rev Re s p ir Dis 1993; 147 : 1557–1561.

78. Sousa AR, Trigg CJ, Lane SJ, e t a l. Effe ct of inhaled glucoc ortic oids on IL-1 beta and IL-1 rec eptor antagonis t (IL-1 ra) ex pres sion in asthmaticbronchial epithe lium. Th o ra x 1997; 52: 407–410.

79. Trigg CJ, Manolitsas ND, Wang J, e t a l. Plac ebo-controlled immunopa-thologic s tudy of four months of inhaled corticos te roids in asthma. AmJ Re s pir Crit Ca re Me d 1994; 150 : 17–22.

80. Wang JH, De valia JL, Xia C, Saps ford RJ, Davie s RJ. Ex pres sion ofRANTES by human bronchial e pithe lial c e lls in vitro and in vivo andthe e ffect of c orticoste roids . Am J Re s p ir Ce ll Mo l Bio l 1996; 14 :27–35.

81. Bai TR, Zhou D, Weir T, e t a l. Substance P (NK1)- and neurokinin A(NK2)-receptor gene ex pre ssion in inflammatory airw ay dise ases . Am JPhy s io l 1995; 269 : L309–L317.

82. Adcock IM, Pete rs M, Ge lder C, Shirasaki H, Brow n CR, Barnes PJ.Inc reased tachykinin re ceptor gene ex pres sion in asthmatic lung and itsmodulation by ste roids . J Mo l End o c rin o l 1993; 11: 1–7.

83. Mak JC, Nishikaw a M, Shirasaki H, Miyayasu K, Barnes PJ. Protec tivee ffects of a glucocorticoid on dow nre gulation of pulmonary beta2-adrene rgic rec eptors in vivo. J Clin Inve s t 1995; 96 : 99–106.

84. Tan S, Hall IP, Dew ar J, Dow E, Lipw orth B. Association betw ee n beta2-adrenoce ptor polymorphism and susc eptibility to bronchodilatordese nsitisation in mode rate ly se ve re stable asthmatic s. La nc e t 1997;350 : 995–999.

85. Colotta F, Re F, Muzio M, e t a l. Interleukin-1 type II rec eptor: a decoytarget for IL-1 that is re gulated by IL-4. Scie nc e 1993; 261 : 472–475.

86. Dow e r SK, Fanslow W, Jacobs C, Waugh S, Sims JE, Widme r MB.Interleukin-I antagonists. The r Im m u n o l 1994; 1: 113–122.

87. Re F, Muzio M, De Rossi M, e t a l.The type II ‘rece ptor’ as a decoy targetfor interleukin 1 in polymorphonucle ar leukoc ytes: charac te rization ofinduction by dex ame thasone and ligand binding propertie s of there leased de coy rece ptor. J Exp Me d 1994; 179 : 739–743.

88. Levine SJ, Logun C, Chopra DP, Rhim JS, She lhamer JH. Prote in kinaseC, interleukin-1 beta, and corticosteroids regulate she dding of the typeI, 55 kDa TNF re ceptor from human airw ay epithe lial ce lls. Am J Re s p irCe ll Mo l Bio l 1996; 14: 254–261.

89. Geiste rfe r M, Richards C, Baumann M, Fe y G, Gyw nne D, Gauldie J.Re gulation of IL-6 and the hepatic IL-6 rec eptor in acute inflammationin vivo. Cy to k ine 1993; 5: 1–7.

90. Snyers L, De Wit L, Conte nt J. Glucocortic oid up-regulation of high-affinity interleukin 6 receptors on human epithe lial c e lls. Pro c Na tlAc a d Sci USA 1990; 87 : 2838–2842.

91. Takizaw a H, Ohtoshi T, Yamashita N, Oka T, Ito K. Interleukin 6-rece ptorex press ion on human bronchial epithe lial c e lls: regulation by IL-1 andIL-6. Am J Phy s io l 1996; 270 : L346–L352.

92. Guo FH, De Rae ve HR, Rice TW, Stuehr DJ, Thunnis sen FB, Erzurum SC.Continuous nitric ox ide synthes is by inducible nitric ox ide synthase innormal human airw ay epithe lium in v ivo. Pro c Na tl Ac a d Sci USA 1995;92: 7809–7813.

93. Robbins RA, Barnes PJ, Springall DR, e t a l. Ex press ion of induciblenitric ox ide in human lung e pithe lial ce lls. Bio ch em Bio phy s Re sCo m m u n 1994; 203 : 209–218.

94. Berkman N, Robichaud A, Robbins RA, e t a l. Inhibition of induciblenitric ox ide synthase ex pre ssion by interle ukin-4 and interleukin-13 inhuman lung epithe lial ce lls. Im m u n o lo g y 1996; 89: 363–367.

95. Adcock IM, Brow n CR, Kw on O, Barne s PJ. Ox idative stre s s induce s NFkappa B DNA binding and induc ible NOS mRNA in human epithe lialce lls. Bio ch em Bio phy s Re s Co m m u n 1994; 199 : 1518–1524.

96. Xie QW, Kashiw abara Y, Nathan C. Role of transcription factor NF-kappaB/Rel in induc tion of nitr ic ox ide synthase . J Bio l Ch em 1994; 269 :4705–4708.

97. Barne s PJ. Nitric ox ide and asthma. Re s Im m u n o l 1995; 146 :698–702.

98. Schalkw ijk C, Vervoorde ldonk M, Pfe ils chifte r J, Marki F, van denBosch H. Cytokine - and forskolin-induced synthes is of group IIphospholipase A2 and pros taglandin E2 in rat mesangial c e lls ispre ve nted by dex amethasone . Bio c he m Bio phy s Re s Co m m u n 1991;180 : 46–52.

99. Mitchell JA, Be lvisi MG, Akarasereenont P, e t a l. Induction of cyclo-ox ygenase-2 by cytokines in human pulmonary epithe lial c e lls :re gulation by dex ame thasone . Br J Pha rm a co l 1994; 113 :1008–1014.

100. van der Ve lden VHJ, Nabe r BAE, Van de r Spoe l P, Hoogste de n HC,Versne l MA. Cytokines and gluc ocorticoids modulate human bronchialepithe lial ce ll pe ptidases. Cy to k in e 1998; 10: 55–65.

101. Lang Z, Murlas CG. Dex amethasone increase s airw ay epithe lial ce llneutral endopeptidase by enhancing transcription and new prote insynthes is. Lu n g 1993; 171 : 161–172.

102. Borson DB, Gruene rt DC. Glucocorticoids induc e neutral endopepti-dase in transforme d human tracheal epithe lial ce lls . Am J Phy s io l 1991;260 : L83–L89.

103. Nadel JA, Borson DB. Modulation of neuroge nic inf lammation byneutral e ndopeptidase . Am Rev Re s pir Dis 1991; 143 : S33–S36.

104. Sont JK, van Krieken JH, van Klink HC, e t a l. Enhanced ex press ion ofneutral endopeptidase (NEP) in airw ay epithe lium in biops ie s fromsteroid- versus nonste roid-tre ated patients w ith atopic asthma. Am JRe s pir Ce ll Mo l Bio l 1997; 16: 549–556.

105. Gandhi CR, Be rkow itz DE, Watkins WD. Endothe lins. Biochemistry andpathophysiolog ic actions. An e s th e s io lo g y 1994; 80: 892–905.

106. Vittori E, Marini M, Fasoli A, De Franchis R, Mattoli S. Inc reasedex press ion of endothe lin in bronchial epithe lial ce lls of asthmaticpatients and e ffect of c orticoste roids . Am Rev Re s pir Dis 1992; 146 :1320–1325.

107. Hay DW, Hubbard WC, Undem BJ. Endothe lin-induce d c ontrac tion andmediator re lease in human bronchus. Br J Pha rm a c o l 1993; 110 :392–398.

108. Mattoli S, Mezzetti M, Riva G, Allegra L, Fasoli A. Specific binding ofendothe lin on human bronchial smooth musc le ce lls in culture andse cre tion of endothe lin-like material from bronchial epithe lial ce lls . AmJ Re s pir Ce ll Mo l Bio l 1990; 3 : 145–151.

109. Barne s PJ. Endothe lins and pulmonary diseases . J App l Phy s io l 1994;77: 1051–1059.

110. Mansoor AM, Honda M, Saida K, e t a l. Endothe lin induc ed collage nre modeling in ex perimental pulmonary hypertension. Bio ch em Bio -phy s Re s Co m m u n 1995; 215 : 981–986.

111. Rizvi MA, Katw a L, Spadone DP, Myers PR. The e ffects of endothe lin-1 oncollage n type I and type III synthe sis in c ultured porc ine coronary arte ryvasc ular smooth musc le ce lls . J Mo l Ce ll Bio l 1996; 28 : 243 –252.

112. Daw es KE, Cambrey AD, Campa JS, e t a l. Change s in collagenmetabolism in response to endothe lin-1: ev idence for fibroblasthe te roge ne ity. In t J Bio c hem Ce ll Bio l 1996; 28: 229 –238.

113. Springall DR, How arth PH, Counihan H, Djukanovic R, Holgate ST,Polak JM. Endothe lin immunore activity of airw ay epithe lium inasthmatic patients . La nc e t 1991; 337 : 697–701.

114. Redington AE, Springall DR, Meng QH, e t a l. Immunore active e ndothe-lin in bronchial biopsy spec imens : increase d ex press ion in asthma andmodulation by c or tic oste roid therapy. J Alle rg y Clin Im m u n o l 1997;100 : 544–552.

115. Redington AE, Springall DR, Meng QH, e t a l. Immunore active e ndothe-lin in bronchial biopsy spec imens : increase d ex press ion in asthma andmodulation by c or tic oste roid therapy. J Alle rg y Clin Im m u n o l 1997;100 : 544–552.

116. Cronste in BN, Kimmel SC, Levin RI, Martiniuk F, We issmann G. Amechanism for the antiinflammatory e ffects of c orticoste roids: theglucoc ortic oid rece ptor regulate s leukocyte adhes ion to e ndothe lialce lls and ex pres sion of endothe lial-leukoc yte adhes ion molec ule 1 andinterce llular adhesion molec ule 1. Pro c Na tl Aca d Sci USA 1992; 89 :9991–9995.

117. van de Stolpe A, Caldenhove n E, Raaijmakers JA, van der Saag PT,Koende rman L. Glucocortic oid-mediated repre ss ion of interce llularadhes ion mole cule -1 ex press ion in human monocytic and bronchialepithe lial ce ll lines . Am J Re s p ir Ce ll Mo l Bio l 1993; 8: 340–347.

118. Paolie r i F, Battifora M, Ricc io AM, e t a l. Inhibition of adhe sion moleculesby budesonide on a human e pithe lial ce ll line (lung c arc inoma). Alle rg y1997; 52: 935 –943.

119. Montefort S, Herbert CA, Robinson C, Holgate ST. The bronchialepithe lium as a target for inflammatory attack in asthma. [Re view ]. ClinExp Alle rg y 1992; 22 : 511–520.

120. Sato M, Takizaw a H, Kohyama T, e t a l. Eosinophil adhesion to humanbronchial epithe lial ce lls : re gulation by cytokine s. In t Arc h Alle rg yIm m u no l 1997; 113 : 203–205.

121. Goulding NJ, Guyre PM. Regulation of inf lammation by lipocortin 1.Im m u no l To da y 1992; 13 : 295–297.

122. Geisow MJ, Walker JH, Bouste ad C, Taylor W. Annex ins –new family ofCa2+-regulated-phospholip id binding prote in. Bio s ci Rep 1987; 7 :289–298.

123. Flow er RJ. Eleventh Gaddum me morial lec ture . Lipocortin and themechanism of ac tion of the glucocorticoids . Br J Pha rm a co l 1988; 94 :987–1015.

124. Wallner BP, Mattaliano RJ, Hession C, e t a l. Cloning and ex pre ssion ofhuman lipocortin, a phospholipase A2 inhibitor w ith potential anti-inflammatory activ ity. Na tu re 1986; 320 : 77–81.

125. Ve rheggen MM, De Bont HI, Adriaansen-Soeting PWC, e t a l. Ex press ionof lipocortins in human bronchial epithe lial c e lls: e ffe cts of IL-1B, TNF-a, LPS and dex ame thasone . Me d In fla m 1996; 5: 210–217.



126. van Hal PT, Ove rbe ek SE, Hoogsteden HC, e t a l. Eicos anoids andlipocortin-1 in BAL fluid in asthma: e ffects of smoking and inhale dglucoc ortic oids. J App l Phy s io l 1996; 81: 548–555.

127. Levine SJ, Benfie ld T, She lhamer JH. Corticos te roids induc e intrace llularinterleukin-1 rece ptor antagonist type I ex pres sion by a human airw ayepithe lial ce ll line . Am J Re s pir Ce ll Mo l Bio l 1996; 15: 245–251.

128. Abbinante-Nis sen JM, Simpson LG, Leikauf GD. Corticosteroids increasese cre tory le ukocyte prote ase inhibitor transcript leve ls in airw ayepithe lial ce lls. Am J Phy s io l 1995; 268 : L601–L606.

129. Djukanovic R, Home yard S, Gratziou C, e t a l. The e ffect of tre atmentw ith oral cortic os te roids on asthma symptoms and airw ay inf lamma-tion. Am J Re s pir Crit Ca re Me d 1997; 155 : 826–832.

130. Naseer T, Minshall EM, Leung DY, e t a l. Ex pre ss ion of IL-12 and IL-13mRNA in asthma and the ir modulation in response to ste roid therapy.Am J Re s p ir Cr it Ca re Med 1997; 155 : 845–851.

131. Robinson D, Hamid Q, Bentle y A, Ying S, Kay AB, Durham SR. Ac tivationof CD4+ T ce lls, increase d TH2-type c ytokine mRNA ex pre ssion, andeosinophil re cruitment in bronchoalveolar lavage afte r alle rge n inhala-tion challe nge in patie nts w ith atopic asthma. J Alle rg y Clin Im m u n o l1993; 92: 313 –324.

132. Bentle y AM, Hamid Q, Robinson DS, e t a l. Prednisolone treatment inasthma. Re duc tion in the numbers of eosinophils , T ce lls, tryptase-onlypositive mast ce lls, and modulation of IL-4, IL-5, and interfe ron-gammacytokine gene ex pre ssion w ithin the bronchial mucosa. Am J Re s pirCrit Ca re Me d 1996; 153 : 551–556.

133. Djukanovic R, Wilson JW, Britte n KM, e t a l. Effect of an inhaledcorticos te roid on airw ay inf lammation and symptoms in asthma. AmRe v Re s p ir Dis 1992; 145 : 669–674.

134. Booth H, Richmond I, Ward C, Gardiner PV, Harkaw at R, Walters EH.Effect of high dose inhaled fluticasone propionate on airw ay inf lamma-tion in asthma. Am J Re s pir Crit Ca re Med 1995; 152 : 45–52.

135. Olivie ri D, Chetta A, De l Donno M, e t a l. Effect of short-te rm treatmentw ith low -dose inhale d f luticasone propionate on airw ay inflammationand re modeling in mild asthma: a p lacebo-c ontrolle d s tudy. Am J Re s p irCrit Ca re Me d 1997; 155 : 1864–1871.

136. Duddridge M, Ward C, Hendrick DJ, Walters EH. Change s in bronchoal-veolar lavage inf lammatory c e lls in asthmatic patients tre ated w ith highdose inhale d bec lomethasone dipropionate . Eu r Re s pir J 1993; 6:489–497.

137. Schle ime r RP, Schulman ES, MacGlashan DW Jr, e t a l. Effe cts ofdex amethasone on me diator re le ase from human lung fragments andpurified human lung mast ce lls. J Clin Inve s t 1983; 71: 1830–1835.

138. Schle ime r RP. Effec ts of gluc oc orticoste roids on inflammatory ce llsre le vant to the ir therape utic applications in asthma. Am Rev Re s p ir Dis1990; 141 : S59–S69.

139. Davis MM, Bjorkman PJ. T-ce ll antigen rec eptor gene s and T-ce llre cognition. Na ture 1988; 334 : 395–402.

140. Inaba K, Ste inman RM. Res ting and sens itize d T lymphocytes ex hibitdis tinct s timulatory (antigen-presenting ce ll) require me nts for grow thand lymphokine re lease . J Ex p Med 1984; 160 : 1717–1735.

141. Molle r GM, Overbeek SE, van He lde n-Meeuw sen CG, e t a l. Inc reasednumber of dendritic ce lls in the bronchial mucosa of atopic asthmatics:dow nre gulation by inhaled corticos te roids. Clin Ex p Alle rg y 1996; 26 :517–524.

142. Bousquet J, Chanez P, Lacoste JY, e t a l. eosinophilic inf lammation inasthma. N Eng l J Med 1990; 323 : 1033–1039.

143. Wardlaw AJ, Kay AB. The role of the eosinophil in the pathogene sis ofasthma. Alle rg y 1987; 42: 321–335.

144. Djukanovic R, Wilson JW, Britte n KM, e t a l. Quantitation of mast c e llsand e os inophils in the bronchial mucosa of symptomatic atopicasthmatics and healthy c ontrol subje cts using immunohistoche mis try.Am Rev Re s pir Dis 1990; 142 : 863–871.

145. Djukanovic R, Roche WR, Wilson JW, e t a l. Mucosal inf lammation inasthma. Am Rev Re s pir Dis 1990; 142 : 434–457.

146. Azzaw i M, Bradle y B, Je ffe ry PK, e t a l. Identific ation of activate d Tlymphocyte s and eosinophils in bronchial biopsie s in stable atopicasthma. Am Rev Re s pir Dis 1990; 142 : 1407–1413.

147. Beas le y R, Roche WR, Roberts JA, Holgate ST. Cellular events in thebronchi in mild asthma and afte r bronchial provoc ation. Am Re v Re s p irDis 1989; 139 : 806 –817.

148. Adelroth E, Rosenhall L, Johansson SA, Linden M, Ve nge P. Inflammatoryce lls and e os inophilic activity in asthmatic s investigate d by bronchoal-veolar lavage . The e ffec ts of antiasthmatic tre atment w ith budesonideor te rbutaline . Am Rev Re s p ir Dis 1990; 142 : 91–99.

149. Lope z AF, Williamson DJ, Gamble JR, e t a l. Recombinant humangranulocyte-macrophage c olony-stimulating factor s timulates in vitromature human neutrophil and eosinophil function, surface rec eptorex press ion, and surv ival. J Clin Inve s t 1986; 78: 1220 –1228.

150. Ow en WF, Rothenberg ME, Pete rsen J, e t a l. Inte rleukin 5 andphe notypic ally alte re d eosinophils in the blood of patie nts w ith theidiopathic hype reos inophilic syndrome . J Exp Med 1989; 170 :343–348.

151. Rothe nberg ME, Ow en WF Jr, Silberste in DS, e t a l. Human eosinophilshave prolonge d surv ival, enhanced func tional prope rtie s, and becomehypode nse w he n ex posed to human inte rleukin 3. J Clin Inve s t 1988;81: 1986–1992.

152. Clutterbuck EJ, Hirst EM, Sanderson CJ. Human interleukin-5 (IL-5)re gulate s the production of eosinophils in human bone marrowculture s: comparison and inte raction w ith IL-1, IL-3, IL-6, and GMCSF.Blo o d 1989; 73: 1504 –1512.

153. Nakamura Y, Azuma M, Okano Y, e t a l. Upregulatory e ffects ofinterleukin-4 and interle ukin-13 but not interleukin-10 on granulocyte /macrophage colony-stimulating factor produc tion by human bronchialepithe lial ce lls. Am J Re s pir Ce ll Mo l Bio l 1996; 15: 680–687.

154. Robinson DS, Assoufi B, Durham SR, Kay AB. Eosinophil c ationicprote in (ECP) and eosinophil prote in X (EPX) conc entrations in serumand bronchial lavage fluid in asthma. Effec t of prednisolone tre atment.Clin Exp Alle rg y 1995; 25 : 1118–1127.

155. Laitine n LA, Laitinen A, Haahte la T. A comparative study of the e ffec ts ofan inhaled corticosteroid, budesonide , and a beta 2-agonist, te rbutaline ,on airw ay inf lammation in new ly diagnose d asthma: a randomized,double-blind, paralle l-group controlled trial. J Alle rg y Clin Im m u n o l1992; 90: 32–42.

156. Keatings VM, Jatakanon A, Worsde ll YM, Barnes PJ. Effects of inhaled andoral gluc ocorticoids on inflammatory indices in asthma and COPD. AmJ Re s pir Crit Ca re Me d 1997; 155 : 542–548.

157. Burke C, Pow e r CK, Norris A, Condez A, Schmeke l B, Poulter LW. Lungfunction and immunopathologic al change s afte r inhale d c or ticoste roidtherap y in asthma. Eu r Re s pir J 1992; 5: 73–79.

158. Krouw e ls FH, van de r He ijden JF, Lutter R, van Ne erven RJ, Janse n HM,Out TA. Glucocortic os te roids affect functions of airw ay- and blood-derived human T-c e ll c lones, favoring the Th1 profile through tw omechanisms. Am J Re s p ir Ce ll Mo l Bio l 1996; 14: 388–397.

159. Lundgren R. Scanning e lec tron microscop ic studies of bronchialmucosa be fore and during treatment w ith bec lome thasone dipropio-nate inhalations . Sca nd J Re s pir Dis Supp l 1977; 101 : 179–187.

160. Lundgren JD, Hirata F, Marom Z, e t a l. Dex amethasone inhibitsre spiratory glycoc onjugate se cre tion from fe line airw ays in vitro by theinduction of lipoc ortin (lipomodulin) synthesis. Am Rev Re s pir Dis1988; 137 : 353–357.

161. Je ffe ry PK, Godfre y RW, Ade lroth E, Nelson F, Rogers A, Johansson SA.Effects of tre atment on airw ay inf lammation and th ickening ofbase ment membrane retic ular collagen in asthma. A quantitative lightand e lectron mic rosc op ic study. Am Rev Re s p ir Dis 1992; 145 :890–899.

162. Schle ime r RP. An overview of glucocortic oid anti-inflammatory ac tions .Eur J Clin Pha rm a co l 1993; 45 : S3 –S7.

163. van de Graaf EA, Out TA, Roos CM, Janse n HM. Re spiratory me mbranepermeability and bronchial hyperreactiv ity in patients w ith stableasthma. Effe cts of therapy w ith inhale d s te roids. Am Rev Re s pir Dis1991; 143 : 362–368.

164. Erje falt I, Pe rsson CG. Anti-asthma drugs attenuate inflammatoryleakage of p lasma into airw ay lumen. Acta Phy s io l Sca n d 1986; 128 :653–654.

165. Boschetto P, Rogers DF, Fabbri LM, Barne s PJ. Cortic oste roid inhibitionof airw ay mic rovascular leakage. Am Rev Re s p ir Dis 1991; 143 :605–609.

166. Shimura S, Sasaki T, Ikeda K, Yamauchi K, Sasaki H, Takishima T. Directinhibitory action of glucocorticoid on glyc oconjugate se cre tion fromairw ay submuc osal glands. Am Rev Re s p ir Dis 1990; 141 :1044–1049.

167. Osterman K, Carlholm M, Eke lund J, e t a l. Effec t of 1 year dailytreatment w ith 400 microg budesonide (Pulmicort Turbuhale r) innew ly diagnosed asthmatics. Eu r Re s p ir J 1997; 10: 2210–2215.

168. Simons FE. A comparison of bec lome thasone , salmete rol, and place boin children w ith asthma. Canadian Bec lome thasone Dipropionate-Salmete rol Xinafoate Study Group. N En g l J Med 1997; 337 :1659–1665.

169. Barne s PJ. Effect of c ortic oste roids on airw ay hyperre sponsivene ss. AmRe v Re s p ir Dis 1990; 141 : S70–S76.

170. Cockcroft DW, Murdock KY. comparative e ffe cts of inhaled salbutamol,sodium cromoglycate , and be c lomethasone dipropionate on alle rgen-induced early asthmatic re sponses , late asthmatic responses , andincreased bronchial re sponsivene ss to his tamine . J Alle rg y ClinIm m u no l 1987; 79: 734–740.

171. Dahl R, Johansson SA. Importance of duration of treatment w ith inhale dbudesonide on the immediate and late bronchial reaction. Eu r J Re s pirDis Su ppl 1982; 122 : 167–175.

172. Burge PS. The e ffects of corticos te roids on the immediate asthmaticre action. Eu r J Re s p ir Dis Su pp l 1982; 122 : 163–166.

173. Martin GL, Atkins PC, Dunsky EH, Zw eiman B. Effec ts of theophylline ,te rbutaline , and prednisone on antigen-induce d bronchospasm andmediator re le ase . J Alle rg y Clin Im m u no l 1980; 66: 204–212.

ACKNOWLEDGEMENTS. I acknow le dge Mr T. M. van Os for preparing thefigure .

Received 28 April 1998;accepted 7 May 1998



Immunology 1994 82 192-199

Topical glucocorticosteroid (fluticasone propionate) inhibits cellsexpressing cytokine mRNA for interleukin-4 in the nasal mucosa in

allergen-induced rhinitis

K. MASUYAMA,* M. R. JACOBSON,* S. RAK,*t Q. MENG,* R. M. SUDDERICK,* A. B. KAY,*0. LOWHAGENt Q. HAMID*j & S. R. DURHAM* *Department of Allergy & Clinical Immunology, National Heart

& Lung Institute, London, U.K. and tSahlgren Hospital, Gothenburg, Sweden

SUMMARY

Allergen-induced late nasal responses are associated with recruitment and activation of Tlymphocytes and eosinophils and preferential mRNA expression for T-helper type 2 (Th2)cytokines. We tested the hypothesis that topical corticosteroids may inhibit late responses byinhibiting cells expressing mRNA for Th2 cytokines. A randomized double-blind placebo-controlled trial of topical corticosteroid (fluticasone propionate) was performed in 48 adult grass

pollen-sensitive patients. Nasal biopsies were taken at baseline and repeated 24 hr after local nasalallergen provocation following 6 weeks treatment with either fluticasone propionate 200 Pig or

placebo nasal spray twice daily. Baseline mRNA expression for interleukin-4 (IL-4) (P = 0 01) andIL-5 (P = 0 002) was higher in the patients than in normal controls. Topical corticosteroidtreatment significantly inhibited immediate nasal symptoms, with almost complete inhibition ofthe late response following allergen challenge. This was associated with a marked decrease in theallergen-induced increases in cells expressing mRNA for IL-4 (P = 0 002) but not for IL-5.Inhibition of the late response was also accompanied by decreases in CD25 + cells, presumed Tlymphocytes and eosinophils. A significant correlation was observed between the decreases in IL-4mRNA+ cells and in eosinophils after treatment (r = 0 46, P < 0 05). These results suggest thatprolonged treatment with topical corticosteroid inhibits allergen-induced early and late nasalresponses and the associated tissue eosinophilia, and that, at least in part, this may result frominhibition of cells expressing mRNA for IL-4.

INTRODUCTION

Mosmann et al. l have described two types of murine T-lymphocyte clones according to their profile ofcytokine release.T-helper cell type 1 cells produce predominantly interleukin-2(IL-2) and interferon-y (IFN-y), whereas Th2-type cells expressmainly IL-4 and IL-5. Recent studies have suggested theinvolvement of Th2-type cytokines in atopic allergic diseases.For example, IL-5 promotes the differentiation, prolongssurvival, and selectively enhances vascular adhesion ofeosinophils.2-5 IL-4 induces isotype switching of B cells infavour of IgE synthesis.6 Human allergen-specific T-cell clonesderived from atopic donors have been shown to exhibit a Th2-type cytokine profile following allergen stimulation.7'8Although Th2-type cytokines were originally described asproducts ofT lymphocytes, more recent studies have identified

Received 23 November 1993; revised 30 December 1993; accepted 1February 1994.

tPresent address: Meakins Christie Laboratories, McGill Univer-sity, 3626 St Urbain St., Montreal, Quebec, Canada H2X 2PZ.

Correspondence: Dr S. R. Durham, Dept. of Allergy & ClinicalImmunology, National Heart & Lung Institute, Dovehouse Street,London SW3 6LY, U.K.

human mast cells,9 basophils10 and eosinophils as alternativesources. 11'12 We recently performed in situ hybridization studiesof biopsies of human skin and nose obtained during allergen-induced late responses.13'14 At 24 hr after allergen there weresignificant increases in cells expressing cytokine mRNA forIL-4, IL-5 and granulocyte-macrophage colony-stimulatingfactor (GM-CSF). These observations support the view thatTh2-type mRNA expression may be implicated in thedevelopment of late-phase responses.

Topical glucocorticosteroids are effective in the clinicalmanagement of allergic rhinitis and inhibit both early and latenasal responses and the associated eosinophilia'5 followingallergen challenge, although their mode of action remainsunknown. In this study we have examined the influence of anovel potent topical glucocorticosteroid (fluticasone propio-nate)'6 on early- and late-phase nasal responses, and cells andcytokine mRNA expression in the nasal mucosa in biopsiesobtained 24 hr after allergen challenge.

MATERIALS AND METHODS

PatientsForty-eight non-smoking Timothy grass pollen-sensitive

192

Glucocorticosteroid inhibition of IL-4 expression

patients were recruited from the out-patient clinic of theAsthma and Allergy Centre, Department of Medicine, SahlgrenHospital, Gothenburg, Sweden. They were selected on the basisof (1) a clinical history of moderate to severe hayfever for atleast 2 years; and (2) a positive skin prick test (wheal diametergreater than 5 mm) to Timothy grass pollen extract (Phleumpratense; Soluprick; ALK, Horsholm, Denmark). Patients wereexcluded if they gave a history of perennial allergy, birch pollenallergy, had previously received immunotherapy or were takingregular topical or oral medication.

Study designThe study was performed with the approval of the EthicsCommittees of The Royal Brompton National Heart and LungHospital, London, U.K. and Sahlgren Hospital, Gothenburg,Sweden. All patients gave informed written consent. The 48patients were stratified according to disease severity into'moderate' and 'severe' groups on the basis of the clinicalhistory and skin prick test results prior to randomization. Onefemale patient dropped out because of pregnancy. Patientswere randomized in double-blind fashion to receive 6 weekstreatment with either aqueous fluticasone propionatel6(Flixonase, Glaxo, Sweden), two sprays (200 yg), or a matchedplacebo spray (containing the aqueous diluent) twice daily.Baseline nasal biopsies were performed out of season in allpatients at a time when they were asymptomatic. None of thepatients had received topical corticosteroids in the previous 6months or had had an upper respiratory infection in the monthprior to biopsy. After 6 weeks treatment all patients underwentlocal nasal provocation with a grass pollen extract followed bya second nasal biopsy at 24 hr after allergen.

Blood samplesVenous blood was taken for measurement of total and allergen-specific IgE concentrations by ELISA.

Nasal provocationNasal provocation was performed by the application of a 4-mmfilter paper disk presoaked with 7 u1 solution containing 1000biological units (BU) of Timothy grass pollen extract to theundersurface of the inferior turbinate, 2 cm distal to its anteriorinsertion. The paper disk was held in position by an attached finecotton thread which was taped to the cheek. The disk was

removed after 10 min.Nasal symptoms were recorded as number of sneezes,

weight of nasal secretions and degree of nostril blockagebetween 0-30 min, 30-60 min, 1-4 hr, 4-10 hr and 10-24 hr.Data were summed for the early response (0-60 min) and lateresponse (1-24 hr). Sneezes were recorded as the total numberduring a given interval. The patients recorded the degree ofnostril blockage by gently applying the thumb to the oppositenostril without deforming the nose and inhaling and exhalinggently through the nose. The degree of blocking was recordedas 'no blocking' to 'complete blockage' on a scale of 0-4. Nasalsecretions were collected by blowing into paper tissues whichwere supplied in preweighted resealable plastic bags. Bags were

reweighed after 24 hr to determine the weight of nasalsecretions. Secretions were expressed as weight in grams.

Nasal biopsyNasal biopsies were obtained at baseline and at 24 hr after

allergen challenge. Local anaesthesia was achieved by applyinga small cotton wool plug soaked in 3% cocaine and 0-025%adrenaline immediately below the inferior turbinate for 10 min.A 2-5-mm biopsy was taken using Gerritsma forceps.'7 Anylocal bleeding was controlled by cautery using a silver nitratestick. After biopsy patients were kept under medical obser-vation for 1 hr.

In situ hybridizationRiboprobes, both antisense (complementary RNA) andsense (having an identical sequence to mRNA), were preparedfrom cDNA for IL-2, IL-4, IL-5 and IFN-y as describedpreviously.'3"18"19 cDNA were inserted into different pGEMvectors and linearized with appropriate enzymes beforetranscription. Transcription was performed in the presence of[35S]uridine triphosphate ([35S]UTP) and the appropriate T7 or

SP6 RNA polymerases. For in situ hybridization, 10-pMcryostat sections were processed on poly-L-lysine-coatedslides. Cryostat sections were permeabilized with 0 3%solution of Triton X-100 in phosphate-buffered saline (PBS)for 15 min and then with proteinase K (1 pg/ml) in 0-1 M Triscontaining 50mM EDTA (pH 8) for 15 min at 37°. The reactionwas terminated by immersion of the preparations in 4%paraformaldehyde for 5 min. In order to inhibit non-specificbinding of 35S, the preparations were treated with 10 mmiodoacetamide and 10 mM N-ethylmaleimide for 30 min, 0 5%acetic anhybride in 0-1 M triethanolamine for 10 min, and 0 1 M

triethanolamine for 2 min. Prehybridization was carried outwith 50% formamide and 2 x standard saline citrate (SSC) for30 min at 400. For hybridization, antisense or sense probes(0 5 x 106 c.p.m./section) diluted in hybridization buffer con-

taining 100 mm dithiothrieol (DTT)15 were used. Hybridizationwas performed in a humid chamber for 16 hr at 42°.Posthybridization washing was performed in a decreasingconcentration of SSC (4 x SSC-0 1 x SSC) at 45°. Unhybrid-ized single-stranded RNA were removed by treating thepreparations with a 2 x SSC solution containing RNase A(20 pg/ml) for 30 min at 45°. After dehydration, the sectionswere immersed in K-5 emulsion (Ilford, U.K.), and exposed for10 days. The autoradiographs were developed in Kodak D-19and counterstained with haematoxylin.

Positive controls for IL-2, IL-4 and IL-5 were cytospinsprepared from a peripheral blood T-lymphocyte clone obtainedfrom a patient with the hyper-IgE syndrome. This clone was

treated with concanavalin A and, after stimulation, expressedmRNA for IL-2, IL-4 and IL-5. Phytohaemagglutin in (PHA)-stimulated blood mononuclear cells were used for the IFN-ycontrols. For the negative controls nasal biopsies were hybridizedusing sense probes for the relevant cytokines. In addition,sections were treated with RNase A solution before theprehybridization step and then hybridized with antisense probes.

ImmunohistologyImmunohistology was performed on 6-,um cryostat sections(fixed for 7 min in 60: 40 acetone: methanol) using the alkalinephosphatase-anti-alkaline phosphatase method, as describedpreviously. 16 The monoclonal antibodies used were CD4(T-helper lymphocytes), CD25 (IL-2-receptor-positive cells)(both from Becton Dickinson, Cowley, U.K.) and EG2('activated', eosinophil cationic protein-secreting eosinophils;Kabia Pharmacia, Milton Keynes, U.K.).

193

K. Masuyama et al.

QuantificationFor in situ hybridization and immunohistology sections were

counted 'blind' in coded random order using an Olympusmicroscope with an eyepiece graticule at 200 x magnification.The graticule (0-202 mm2) was orientated beneath the epithelialbasement membrane and counts were made along the wholelength of the biopsy. The depth counted (0A45 mm) included thewhole lamina propia where the majority of inflammatory cellsare located. At least two sections were counted for each markerwhich produced counts from a minimum of six fields.Submucosal counts were expressed as mean counts per highpower field (0-202 mm2; HPF). The within observer coefficientof variation was < 5% for in situ hybridization counts and 8%for immunohistology counts.

Statistical analysisWithin-group comparisons were performed using the Wilcoxonmatched-pairs signed-ranks test. Between-group comparisonswere made using the Mann-Whitney U-test. Correlations were

performed using Spearman's rank method. P values <0.05were considered statistically significant. Statistical tests were

performed using a standard computer package (MinitabRRelease 7; Minitab Inc., State College, PA).

RESULTS

Randomization resulted in 23 patients receiving fluticasone and24 placebo nasal spray. After randomization, three of thefluticasone-treated patients and two placebo-treated patientsdropped out for reasons unconnected to the clinical protocol.Fluticasone- and placebo-treated patients were well matchedfor age, gender, disease severity and for their degree of grass

pollen sensitivity as determined by skin prick tests and allergen-specific IgE antibody concentration (Table 1). Total IgE levelsfor both groups were either normal or moderately increased,and there was no significant difference between the groups.

Six weeks treatment with fluticasone resulted in highlysignificant inhibition of the size of both early and late nasalresponses after grass pollen allergen challenge. There was an

approximate 50% reduction in early sneezes (P < 0-004),degree of nostril blocking (P < 0 0005) and weight of nasalsecretions (P < 0 001) in the actively treated group. During the

Table 1. Clinical data of patients who underwent in situ hybridizationstudies

Treatment

Placebo Fluticasone

Number of patients 22 20Male: female ratio 12:10 11:9Age (years mean ± SD) 32-5 ± 10-4 31-8 i 94RAST* score Timothy grass 3 0 ± 1-1 3-3 i 09

(0-5) (Mean + SD)Total IgE (IU/ml) 45-0 ± (18 7-93-7) 82 5 ± (41-3-197-5)

(median ± interquartileranges)

*Radioallergosorbant test.

25

go

ID20

re- 15

0

10 -

5 -:z

0-

15 -

o 10 -

C,-0

5-

15w

0n0,1 -

-E

6.

3rly response Late response

P=0-004 P<0.00001

P=0.0005 P=0*002

P=0-0008 P=0 00001

Figure 1. Effect of fluticasone on allergen-induced early and late nasalresponses. Number of sneezes, degree of nostril blocking and weight ofsecretions during early (0-60 min) and late (1-24 hr) periods after grasspollen challenge, following 6 weeks treatment with fluticasone (hatchedcolumns) or placebo (open columns) nasal spray. Median values andinterquartile ranges. Mann-Whitney U-test P values are shown.

late phase (1-24 hr) nostril blocking was similarly reduced(P < 0 002) and there was almost complete inhibition of latesneezing (P < 0-0001) and nasal secretions (P < 00001)(Fig. 1).

Nasal biopsies were performed at baseline in all 47 patients(in situ hybridization studies were performed on 42 patients forwhom paired tissue samples were available) and in 12 non-

atopic normal healthy control subjects (Table 2). When thenumber of cells in the nasal submucosa expressing positivehybridization signals for cytokine mRNA was compared, a

small but significant increase in cytokine mRNA for IL-4

Table 2. Number of cells in nasal submucosa/HPF expressing mRNAfor cytokines in grass pollen-sensitive patients compared with normal

healthy controls

Patients ControlsCytokine (n = 42) (n = 12) P value

IL-4 2-1 (2-0, 3.2)* 0 (0, 1-7) 0 01IL-5 2-2 (1-6, 3-0) 0 (0, 1-2) 0 002IFN-y 1 5 (0-2, 2 4) 1 0 (0, 2 7) NSIL-2 1 0 (0, 1 0) 0 5 (0, 4 6) NS

*Values are medians (95% confidence intervals). Comparison wasmade using Mann-Whitney U-test. P values are shown.

NS, not significant.

194

Glucocorticosteroid inhibition ofIL-4 expression

Placebo FluticasoneP=0 00225-

20-

Q.)

+

10-z

E

5-

0-

Before Ag

15 _

10 _

S _

O _PI

I ---I

Before Ag

Figure 2. Effect of fluticasone on the number of cells in the nasalsubmucosa expressing mRNA for IL-4. Data are shown for baseline(before) and at 24 hr after allergen challenge (Ag) following 6 weekstreatment with fluticasone (closed circles) or placebo (open circles).Median values are represented by the horizontal bars. Mediandifferences (Ag minus before) were compared using the Mann-Whitney U-test.

(P = 001) and IL-5 (P = 0002) was observed in rhiniticscompared to controls. In contrast, no differences for IL-2 or

IFN-y were observed. At baseline, when biopsies fromfluticasone- and placebo-treated patients were examined, thenumber of cytokine mRNA+ cells was small (mean counts1-2 5 cells/HPF) and no significant differences for any of thefour cytokines were observed (data not shown).

Within the placebo-treated group, a significant increase inIL-4 mRNA+ cells (P = 0-0001) 24 hr after allergen was

observed. In contrast, in the fluticasone-treated group no

increase in IL-4 was observed. This difference was highlysignificant compared with the placebo group (Fig. 2;P = 0.002). For IL-5, increases were observed 24 hr afterallergen in both placebo (P = 0-0001) and fluticasone(P = 0.0001)-treated patients. Median counts for IL-5mRNA + cells after allergen were reduced in the treatedgroup, although not significantly when compared with theplacebo group (Fig. 3). No changes in cytokine mRNA+ cellsfor IL-2 or IFN-y 24 hr after allergen were observed for eithergroup of patients (Fig. 3). In general, cytokine mRNA+ cellswere identifiable as discrete, well-circumscribed areas of silvergrains (Fig. 4). No hybridization signals were observed whenthe sense probe was employed or following pretreatment ofsections with RNase.

Immunohistology of nasal biopsies from the same patientsobtained 24 hr after allergen revealed fewer numbers of CD4 +T-helper lymphocytes (P = 0-1), CD25 + cells, presumedactivated T lymphocytes (P = 0-03) and EG2+ eosinophils(P = 0'02) in the nasal submucosa in fluticasone-treatedpatients compared with the placebo group (Fig. 5).

In placebo-treated patients, significant correlations were

found at allergen-challenged sites between the number ofCD4+ T lymphocytes and CD25+ cells (r = 0-67, P < 0-01);between CD4+ cells and IL-4 mRNA+ cells (P = 0-45,P < 0 05) and between CD25+ cells and IL-4 mRNA+ cells(r = 0-67, P < 001). Within the fluticasone-treated group

the magnitude of the changes in eosinophil (EG2+)numbers after allergen correlated with changes in IL-4

P=0-56

IL-5

FP

15- P=0*64LL

0.

I

'"; 10 IL-2w _

,+ 5EI44

z

PI FP15 - P=0-26

10 - IFN-y

PI FP

Figure 3. Effect of fluticasone on the number of cells in nasalsubmucosa expressing mRNA for IL-5, IL-2 and IFN-y. Data are

shown for baseline (open bars) and at 24 hr after local allergenprovocation (closed bars) following 6 weeks treatment with placebo (P1)or fluticasone (FP). Results expressed as medians and interquartileranges. Median differences were compared using the Mann-WhitneyU-test.

mRNA+ cells (r = 0-46, P < 0-05), i.e. the lower thenumber of IL-4 mRNA+ cells after fluticasone treatment thelower the number of eosinophils. No significant associationswere found between the numbers of submucosal CD4+ Tlymphocytes, eosinophils, or IL-4 mRNA+ cells at the singletime-point examined and the magnitude of either early or latenasal symptoms for either fluticasone- or placebo-treatedpatients.

DISCUSSION

We have shown that 6 weeks treatment with topicalglucocorticosteroid inhibited early and late nasal responses

and decreased the number of IL-4 mRNA+ cells in nasalbiopsies at 24 hr after local allergen challenge. Glucocortico-steroid treatment also resulted in significantly lower numbers ofCD25 + cells (presumed T lymphocytes)20 and activatedeosinophils at allergen-challenged sites. CD4+ T-lymphocytenumbers were also lower although not significantly whencompared with placebo-treated patients. The magnitude of thedecreases in IL-4 mRNA + cells correlated with the decreases ineosinophils. The findings confirm our previous observationsthat allergen-induced late nasal responses'4'21 and lateresponses occurring in the skin13 and lung22'23 are associatedwith recruitment of activated T lymphocytes and eosinophilsand increased expression of mRNA for Th2-type cytokines.They also support the view that topical corticosteroids may

inhibit in vivo human allergen-induced late-phase responses by

195

christinehurd

Highlight

K. Masuyama et al.

Figure 4. Autoradiographs of cryostat sections of nasal biopsies 24 hr after local allergen provocation from placebo (a) and fluticasone(b)-treated patients. Sections were hybridized with a 35S-labelled antisense IL-4 probe. Autoradiographs of sections of nasal biopsiesfrom a placebo-treated patient hybridized with a 35S-labelled sense IL-4 probe (c), and following treatment with RNase beforehybridization (d).

P=0 P=0.03 P=0-0280- 4 20

60- CD4 3 CD25 15- EG2r~~~~~~~~~~~~

40 2 ~~~101-1020! I

PI FP PI FP PI FP

Figure 5. Cell infiltration (median cell counts ± interquartile ranges) atallergen-challenged sites in nasal biopsies from placebo (open bars) andfluticaseone (closed bars)-treated patients. Numbers of T-helper cells(CD4 +), activated cells (CD25 +) and eosinophils (EG2+) are shown. Pvalues are shown (Mann-Whitney U-test).

suppressing local eosinophilia, and raise the possibility that thisis related to the suppression of IL-4 expression.

The biological properties of Th2-type cytokines arepertinent to the development of late responses. Thus lateresponses are known to be IgE-dependent24-26 and IL-4 mayregulate local antibody production by switching B cells toproduce preferentially IgE. Tissue eosinophilia is characteristicof human late responses.21 IL-5 stimulates eosinophil matura-tion,2 activation,3 and prolongs eosinophil survival in culture5and also, presumably, in tissues. IL-5 also promotes selectivelocal adhesion of eosinophils to human vascular endothelium.4However, the predominant role of Th2-type cytokines duringlate responses, whether in enhancing local IgE or promotingtissue eosinophilia, is unclear.

196

Alm',

Glucocorticosteroid inhibition ofIL-4 expression 197

Corticosteroids are potent inhibitors of late-phaseresponses. Our original hypothesis was that corticosteroidsmight act by suppressing IL-5. In this study, in placebo-treatedpatients, we confirmed our previous observation that lateresponses were accompanied by increases in the number of cellsexpressing IL-4 and IL-5. It turned out that in corticosteroid-treated patients the number of IL-5 mRNA+ cells remainedincreased despite almost complete inhibition of the lateresponse. In contrast, there was a marked inhibition ofallergen-induced increases in IL-4 mRNA+ cells in patientswho had received topical corticosteroid. In addition, cortico-steroid significantly reduced the number of activated eosino-phils in the nasal mucosa.27 There was no close correlationamong cells, cytokines and the magnitude of early or late nasalsymptoms. This is perhaps not surprising since, for ethicalreasons, biopsy could only be performed at one time-point afterallergen. However, it seems likely that the development ofsymptoms may also depend on other influences such aspsychological or neurological factors which we have notexamined. It is of interest that the single patient who hadreceived fluticasone and demonstrated a large increase in IL-4mRNA+ cells also showed a large local eosinophilia andmarked early and late nasal symptoms, in contrast to theremaining steroid-treated patients.A disadvantage of the nasal biopsy technique is that it is not

amenable to the performance of time-course studies. We chosethe 24-hr time-point based on our previous observation ofreadily identifiable changes in T lymphocytes and eosinophils atthis time.21 Thus we may have missed an effect of corticosteroidon cytokine mRNA, including IL-5, at earlier time-points. Forexample, previous studies which have employed nasal lavagehave shown that the peak time of release of mediators ofhypersensitivity28 and the appearance of cells,29 particularlyeosinophils, in nasal washings during late responses may varybetween individuals. In these studies the release of mediatorsincluding histamine, TAME-esterase, prostaglandin D2(PGD2) and bradykinin have been shown to be inhibited bytopical corticosteroids.30

The expression of cytokine mRNA does not equate withprotein release, which was not examined in our study.However, to our knowledge, Th2-type cytokines have notbeen measured in nasal lavage studies during late responses andthe influence of topical steroids on secreted cytokines in vivo isyet to be determined. A further important question is the celltype responsible for the increase in mRNA expression which wehave observed during late responses, and which cell isinfluenced by corticosteroids. Th2 T lymphocytes,7'8 mastcells9 and basophils10'3' are all candidates for IL-4 mRNAexpression in humans. Basophils represent an attractivecandidate since several studies have supported participationof basophils during human late nasal responses.32 Unfortu-nately, there is no specific monoclonal antibody markerdirected against basophils for tissue studies. A very recentstudy which employed immunostaining of cytokines in nasalbiopsies from patients with perennial rhinitis has shown thepresence of IL-4 protein.9 By use of double immunostaining,IL-4 was localized to mast cells using antibodies directedagainst human IL-4 and mast cell tryptase. Our own recentstudies have focused on the cell source of IL-4 mRNA duringlate nasal responses.33 By use of double immunocytochemistry/in situ hybridization of 10-micron cryostat sections, the

majority of IL-4 mRNA+ cells, approximately 80%, were Tlymphocytes, whereas 20% of IL-4 mRNA' cells were mastcells. In contrast, no IL-4 mRNA was co-localized to eithereosinophils or macrophages. In our hands, therefore, theprincipal cellular source of IL-4 transcripts during lateresponses was the T lymphocyte. It therefore seems likely thatT lymphocytes express mRNA for IL-4 but, unlike the mastcell, do not store IL-4 and are therefore not amenable to specificimmunostaining for IL-4 product. The precise cell expressingIL-4 mRNA which is modulated (or inhibited from migratinginto late reaction sites) by corticosteroids has yet to bedetermined.

The influence of glucocorticosteroids on IL-4 production byT lymphocytes in vitro is dependent upon the species of origin.Previous studies showed, rather surprisingly, that IL-4production by murine T lymphocytes in vitro was enhanced.34In contrast, the production of IL-4 and IL-5 by human Tlymphocytes in vitro was inhibitable at both mRNA andprotein levels by corticosteroids.35-37 Our results support theview that corticosteroids also inhibit IL-4, at least at the level ofmRNA expression, in vivo in humans. Furthermore, theinhibition of IL-4 which we observed after allergen wasaccompanied by marked inhibition of both early and latephase allergic responses and the associated tissue eosinophilia.As mentioned above, this effect of corticosteroids may bemediated by inhibition of local IgE-dependent events. How-ever, IL-4 has recently been demonstrated to up-regulateVCAM-1 on human vascular endothelium.38 Eosinophils, butnot neutrophils, adhere to vascular endothelium through VLA-4-VCAM-l-dependent pathways. Thus an attractive alter-native hypothesis may be that corticosteroids inhibit lateresponses by suppressing IL-4-induced up-regulation ofVCAM-1.

In summary, we have shown that prolonged treatment withtopical corticosteroids inhibited early and late nasal responsesand the associated allergen-induced increases in IL-4 mRNA'cells in the nasal mucosa. Corticosteroid treatment was alsoassociated with fewer numbers ofCD25 + cells and eosinophils.It is not possible to determine, in vivo, whether the inhibition ofIL-4 mRNA resulted from corticosteroid-induced inhibition oftranscription of mRNA or simple inhibition of recruitment ofIL-4 mRNA+ cells. The precise cell type involved and thepotential IL-4-dependent pathways influenced by corticoster-oids are yet to be determined.

ACKNOWLEDGMENTSThe study was supported by a grant from the Medical Research CouncilUnited Kingdom and financial assistance from Glaxo Group ResearchLimited. We are grateful to the Glaxo Institute of Molecular BiologyS.A., Geneva, Switzerland, for their kind gifts of IL-2, IL-4 and IFN-yplasmid constructs, to Dr Colin Sanderson for the IL-5 cDNA, and toGlaxo, Sweden for supplying fluticasone and placebo nasal sprays.T-cell clones from the hyper-IgE syndrome were kindly supplied byDr Diana Quint, Glaxo, Greenford, U.K.

REFERENCES

1. MOSMANN T.R., CHERWINSKI H., BOND M.W., GIEDLIN M.A. &COFFMAN R.L. (1986) Two types of murine helper T cell clone. I.

198 K. Masuyama et al.

Definition according to profiles of lymphokine activities andsecreted proteins. J. Immunol. 136, 2348.

2. CLUTrERBUCK E.J., HIRST E.M.A. & SANDERSON C.J. (1989) Humaninterleukin-5 (IL-5) regulates the production of eosinophils inhuman bone marrow cultures: comparison and interaction withIL-1, IL-3, IL-6 and GMCSF. Blood, 73, 1504.

3. LOPEZ A.F., SANDERSON C.J., GAMBLE J.R., CAMPBELL H.D.,YOUNG I.G. & VADAS M.A. (1988) Recombinant humaninterleukin 5 is a selective activator of human eosinophilfunction. J. exp. Med. 167, 219.

4. WALSH G.M., HARTNELL A., WARDLAW A.J., KURIHARA K.,SANDERSON C.J. & KAY A.B. (1990) IL-5 enhances the in vitroadhesion of human eosinophils, but not neutrophils, in a

leucocyte integrin (CD1 1/18)-dependent manner. Immunology, 71,258.

5. ROTHENBERG M.E., PETERSEN J., STEVENS R.L., SILBERSTEIN D.S.,MCKENZIE D.T., AUSTEN K.F. & Owen W.F. (1989) IL-5-dependent conversion of normodense human eosinophils to thehypodense phenotype uses 3T3 fibroblasts for enhanced viability,accelerated hypodensity, and sustained antibody-dependent cyto-toxicity. J. Immunol. 143, 2311.

6. DEL PRETE G.F., MAGGI E., PARRONCHI P., CHRETIEN I., Tim A.,MACCHIA D., Ricci M., BANCHEREAU J., DE VRIES J. &ROMAGNANI S. (1988) IL-4 is an essential factor for the IgEsynthesis induced in vitro by human T cell clones and theirsupernatants. J. Immunol. 140, 4193.

7. WIERENGA E.A., SNOEK M., DE GROOT C., CHRETIEN I., Bos J.D.,JANSEN H.M. & KAPSENBERG M.L. (1990) Evidence for compart-mentalization of functional subsets of CD4 + T lymphocytes inatopic patients. J. Immunol. 144, 4651.

8. PARRONCHI P., MACCHIA D., PICCINNI M.-P., BISWAS P., SIMONELLIC., MAGGI E., Ricci M., ANSARI A.A. & ROMAGNANI S. (1991)Allergen- and bacterial antigen-specific T-cell clones establishedfrom atopic donors show a different profile of cytokine production.Proc. natl. Acad. Sci. U.S.A. 88, 4538.

9. BRADDING P., FEATHER I.H., HOWARTH P.H., MUELLER R., ROBERTSJ.A., BRITTEN K. et al. (1992) Interleukin 4 is localized to andreleased by human mast cells. J. exp. Med. 176, 1381.

10. BRUNNER T., HEUSSER C.H. & DAHINDEN C.A. (1993) Humanperipheral blood basophils primed by interleukin 3 (IL-3) produceIL-4 in response to immunoglobulin E receptor stimulation. J. exp.Med. 177, 605.

11. DESREUMAUX P., JANIN A., COLOMBEL J.F., PRIN L., PLUMAS J.,EMILIE D., TORPIER G., CAPRON A. & Capron M. (1992)Interleukin-5 messenger RNA expression by eosinophils in theintestinal mucosa of patients with coeliac disease. J. exp. Med. 175,293.

12. MOQBEL R., HAMID Q., YING S., BARKANS J., HARTNELL A.,TSICOPOULOS A., WARDLAw A.J. & KAY A.B. (1991) Expression ofmRNA for the granulocyte/macrophage colony-stimulating factor(GM-CSF) in activated human eosinophils. J. exp. Med. 174,749.

13. KAY A.B., YING S., VARNEY V., GAGA M., DURHAM S.R., MOQBELR., WARDLAw A.J. & HAMID Q. (1991) Messenger RNA expressionof the cytokine gene cluster, interleukin 3 (IL-3), IL4, IL-5 andgranulocyte/macrophage colony-stimulating factor, in allergen-induced late-phase cutaneous reactions in atopic subjects. J. exp.Med. 173, 775.

14. DURHAM S.R., YING S., VARNEY V.A., JACOBSON M.R., SUDDERICKR.M., MACKAY I.S., KAY A.B. & HAMID Q.A. (1992) Cytokinemessenger RNA expression for IL-3, IL4, IL-5, and granulocyte/macrophage-colony-stimulating factor in the nasal mucosa afterlocal allergen provocation: relationship to tissue eosinophilia. J.Immunol. 148, 2390.

15. NACLERIO R.M. (1991) Allergic rhinitis. N. Engl. J. Med. 325, 860.16. DOLOVICH J., ANDERSON M., CHODIRKER W., DROUIN M.,

HARGREAVE F.E., HEBERT J., KNIGHT A., O'CONNER M., SMALL P.

& YANG W. (1990) Fluticasone propionate: a large multicentretrial. Respir. Med. 84 (suppi. A), 31.

17. FOKKENS W.J., VROOM Th.-M., GERRITSMA V. & RIJNTJES E. (1988)A biopsy method to obtain high quality specimens of nasal mucosa.Rhinology, 26, 293.

18. HAMID Q., WHARTON J., TERENGHI G., HASSALL C.J.S., AIMI J.,TAYLOR K.M., NAKAZATO H., DIXON J.E., BURNSTOCK G. & POLAKJ.M. (1987) Localization of atrial natriuretic peptide mRNA andimmunoreactivity in the rat heart and human atrial appendage.Proc. natl. Acad. Sci. U.S.A. 84, 6760.

19. HAMID Q.A., BISHOP A.E., SPRINGALL D.R., ADAMS C., GIAID M.A.,DENNY P. et al. (1989) Detection of human probombesin mRNA inneuroendocrine (small cell) carcinoma of the lung: in situhybridization with cRNA probe. Cancer, 63, 266.

20. HAMID Q., BARKANS J., ROBINSON D.S., DURHAM S.R. & KAY A.B.(1992) Co-expression of CD25 and CD3 in atopic allergy andasthma. Immunology, 75, 659.

21. VARNEY V.A., JACOBSON M.R., SUDDERICK R.M., ROBINSON D.S.,IRANI A.-M.A., SCHWARTZ L.B., MACKAY I.S., KAY A.B. &DURHAM S.R. (1992) Immunohistology of the nasal mucosa

following allergen-induced rhinitis: identification of activated Tlymphocytes, eosinophils, and neutrophils. Am. Rev. Respir. Dis.146, 170.

22. ROBINSON D.S., HAMID Q., BENTLEY A., YING S., KAY A.B. &DURHAM S.R. (1993) Activation of CD4+ T cells, increased Th2-type cytokine mRNA expression, and eosinophil recruitment inbronchoalveolar lavage after allergen inhalation challenge in atopicasthmatics. J. Allergy clin. Immunol. (in press).

23. BENTLEY A.M., QIU MENG, ROBINSON D.S., HAMID Q., KAY A.B. &DURHAM S.R. (1993) Increases in activated T lymphocytes,eosinophils, and cytokine mRNA expression for interleukin-5and granulocyte/macrophage colony-stimulating factor in bron-chial biopsies after allergen inhalation challenge in atopicasthmatics. Am. J. Respir. Cell Mol. Biol. 8, 35.

24. DOLOVICH J., HARGREAVE F.E., CHALMERS R., SHIER K.J., GAULDIEJ. & BIENENSTOCK J. (1973) Late cutaneous allergic responses inisolated IgE-dependent reactions. J. Allergy clin. Immunol. 52,38.

25. SOLLEY G.O., GLEICH G.J., JORDON R.E. & SCHROETER AL. (1976)The late phase of the immediate wheal and flare skin reaction: itsdependence upon IgE antibodies. J. clin. Invest. 58, 408.

26. SHAMPAIN M.P., BEHRENS B.L., LARSEN G.L. & HENSON P.M.(1982) An animal model of late pulmonary responses to alternariachallenge. Am. Rev. Respir. Dis. 126, 493.

27. RAK S., JACOBSON M.R., SUDDERICK R.M., MASUYAMA K., KAYA.B., LOWHAGEN 0. & DURHAM S.R. (1993) Effect of topicalfluticasone propionate on allergen-induced early and late nasalresponses and associated tissue eosinophilia. J. Allergy clin.Immunol. 91, 299.

28. NACLERIO R.M., PROUD D., ToGIAs A.G., ADKINSON N.F. MEYERSD.A., KAGEY-SOBOTKA A., PLAUT M., NORMAN P.S. &LICHTENSTEIN L.M. (1985) Inflammatory mediators in lateantigen-induced rhinitis. N. Engl. J. Med. 313, 65.

29. BASCOM R., PIPKORN U., LICHTENSTEIN L.M. & NACLERIO R.M.(1988) The influx of inflammatory cells into nasal washings duringthe late response to antigen challenge: effect of systemic steroidpretreatment. Am. Rev. Respir. Dis. 138, 406.

30. PIPKORN U., PROUD D., LICHTENSTEIN L.M., KAGEY-SOBOTKA A.,NORMAN P.S. & NACLERIO R.M. (1987) Inhibition of mediatorrelease in allergic rhinitis by pretreatment with topical glucocorti-costeroids. N. Engl. J. Med. 316, 1506.

31. PIccINNI M.-P., MACCHIA D., PARRONCHI P., GIUDIZI M.G.,BANI D., ALTERINI R., GROssI A., RICCI M., MAGGI E. &ROMAGNANI S. (1991) Human bone marrow non-B, non-Tcells produce interleukin 4 in response to cross-linkage ofFcE and Fcy receptors. Proc. natl. Acad. Sci. U.S.A. 88,8656.

Glucocorticosteroid inhibition ofIL-4 expression 199

32. BASCOM R., WACHS M., NACLERIO R.M., PIPKORN U., GALLI S.J. &LICHTENSTEIN L.M. (1988) Basophil influx occurs after nasalantigen challenge: effects of topical corticosteroid pretreatment.J. Allergy clin. Immunol. 81, 580.

33. YING S., DuRHAM S.R., JACOBSON M.R., RAK S., MAsuYAMA K.,LOWHAGEN O., KAY A.B. & HAMID Q.A. (1994) T lymphocytes andmast cells express messenger RNA for interleukin-4 in the nasalmucosa in allergen-induced rhintis. Immunology, 82, 200.

34. DAYNES R.A. & ARANEO B.A. (1989) Contrasting effects ofglucocorticoids on the capacity of T cells to produce thegrowth factors interleukin 2 and interleukin 4. Eur. J. Immunol.19, 2319.

35. Wu C.Y., FARGEAS C., NAKAjimA T. & DELESPESSE G. (1991)Glucocorticoids suppress the production of interleukin 4 by humanlymphocytes. Eur. J. Immunol. 21, 2645.

36. BYRON K.A., VARIGoS G. & WooTrON A. (1992) Hydrocrotisoneinhibition of human interleukin-4. Immunology, 77, 624.

37. ROLFE F.G., HUGHES J.M., ARMOUR C.L. & SEWELL W.A. (1992)Inhibition of interleukin-5 gene expression by dexamethasone.Immunology, 77, 494.

38. SCHLEIMER R.P., STERBINSKY S.A., KAISER J., BICKEL C.A., KLUNKD.A., TOMIOKA K. et al. (1992) IL-4 induces adherence of humaneosinophils and basophils but not neutrophils to endothelium:association with expression of VCAM-1. J. Immunol. 148, 1086.

Mechanisms of allergic diseases

Series editors: Joshua A. Boyce, MD, Fred Finkelman, MD, and William T. Shearer, MD, PhD

The biology of the glucocorticoid receptor: New signalingmechanisms in health and disease

Robert H. Oakley, PhD, and John A. Cidlowski, PhD Research Triangle Park, NC

Abbreviations used

ACTH: Adrenocorticotropic hormone

AF: Activation function

AP1: Activator protein 1

b2AR: b2-Adrenergic receptor

CBG: Corticosteroid-binding globulin

DBD: DNA-binding domain

GPCR: G protein–coupled receptor

GR: Glucocorticoid receptor

GRE: Glucocorticoid-responsive element

LBD: Ligand-binding domain

MAPK: Mitogen-activated protein kinase

NF-kB: Nuclear factor kB

nGRE: Negative glucocorticoid-responsive element

NTD: N-terminal transactivation domain

SEGRA: Selective glucocorticoid receptor agonist

Glucocorticoids are primary stress hormones necessary for lifethat regulate numerous physiologic processes in an effort tomaintain homeostasis. Synthetic derivatives of these hormoneshave been mainstays in the clinic for treating inflammatorydiseases, autoimmune disorders, and hematologic cancers. Thephysiologic and pharmacologic actions of glucocorticoids aremediated by the glucocorticoid receptor (GR), a member of thenuclear receptor superfamily of ligand-dependent transcriptionfactors. Ligand-occupied GR induces or represses thetranscription of thousands of genes through direct binding toDNA response elements, physically associating with othertranscription factors, or both. The traditional view thatglucocorticoids act through a single GR protein has changeddramatically with the discovery of a large cohort of receptorisoforms with unique expression, gene-regulatory, andfunctional profiles. These GR subtypes are derived from a singlegene by means of alternative splicing and alternative translationinitiation mechanisms. Posttranslational modification of theseGR isoforms further expands the diversity of glucocorticoidresponses. Here we discuss the origin and molecular propertiesof the GR isoforms and their contribution to the specificity andsensitivity of glucocorticoid signaling in healthy and diseasedtissues. (J Allergy Clin Immunol 2013;132:1033-44.)

Key words: Glucocorticoid receptor, glucocorticoid, isoforms,glucocorticoid signaling

Glucocorticoids are hormones essential for life that aresynthesized and released by the adrenal cortex in a circadianmanner and in response to stress. The secretion of these hormonesis controlled by the hypothalamic-pituitary-adrenal axis (Fig 1).Internal and external signals trigger the hypothalamus to releasecorticotropin-releasing hormone, which acts on the anteriorpituitary to stimulate the synthesis and secretion of

From the Laboratory of Signal Transduction, National Institute of Environmental Health

Sciences, National Institutes of Health, Department of Health and Human Services.

Supported by the Intramural Research Program of the National Institutes of Health/

National Institute of Environmental Health Sciences.

Disclosure of potential conflict of interest: The authors declare that they have no relevant

conflicts of interest.

Received for publication August 1, 2013; revised September 5, 2013; accepted for pub-

lication September 6, 2013.

Available online September 29, 2013.

Corresponding author: John A. Cidlowski, PhD, National Institute of Environmental

Health Sciences, PO Box 12233, MD F3-07, Research Triangle Park, NC 27709.

E-mail: [email protected].

0091-6749

http://dx.doi.org/10.1016/j.jaci.2013.09.007

Terms in boldface and italics are defined in the glossary on page 1034.

adrenocorticotropic hormone (ACTH). ACTH then acts on the ad-renal cortex to stimulate the production and secretion of glucocor-ticoids. Acting on nearly every tissue and organ in the body,glucocorticoids function to maintain homeostasis both inresponse to normal diurnal changes in metabolism and in theface of stressful perturbations. Glucocorticoids regulate aplethora of physiologic processes, including intermediary meta-bolism, immune function, skeletal growth, cardiovascular func-tion, reproduction, and cognition.1,2 In a classic negativefeedback loop, glucocorticoids also target the hypothalamusand anterior pituitary to inhibit the production and release ofcorticotropin-releasing hormone and ACTH and thereby limitboth the magnitude and duration of the glucocorticoid increase(Fig 1).Because of their powerful anti-inflammatory and immunosup-

pressive actions, glucocorticoids are one of the most widelyprescribed drugs in the world today.3,4 Synthetic glucocorticoidshave been indispensable over the last half century for treating in-flammatory and autoimmune diseases, such as asthma, allergy,sepsis, rheumatoid arthritis, ulcerative colitis, and multiple scle-rosis. They are also commonly prescribed to prevent organ trans-plant rejection and to treat cancers of the lymphoid system, suchas leukemias, lymphomas, and myelomas. Unfortunately, thetherapeutic benefits of glucocorticoids are limited by severeside effects that develop in patients chronically treated with thesesteroids.3,5 Adverse effects include osteoporosis, skin atrophy,diabetes, abdominal obesity, glaucoma, growth retardation inchildren, and hypertension. In addition, patients receiving long-term glucocorticoid therapy frequently have tissue-specific gluco-corticoid resistance. Understanding the factors at a molecular

1033

Delta:1_given name

Delta:1_surname


http://dx.doi.org/10.1016/j.jaci.2013.09.007

J ALLERGY CLIN IMMUNOL

NOVEMBER 2013

1034 OAKLEY AND CIDLOWSKI

level that control the cellular response to glucocorticoids is a crit-ical goal of current research because progress in this area willfacilitate the development of novel glucocorticoids withimproved benefit/risk ratios.The physiologic and pharmacologic actions of glucocorticoids

are mediated by the glucocorticoid receptor (GR; NR3C1), amember of the nuclear receptor superfamily of ligand-dependenttranscription factors.6 After glucocorticoid binding, the GR inducesor represses the transcription of target genes, which can compriseup to 10% to 20% of the human genome.7-9 Consistent with thepleiotropic actions of glucocorticoids, the GR is expressed in nearlyevery cell of the body and is necessary for life after birth.10 Thecellular response to glucocorticoids is remarkably diverse, exhibit-ing profound variability in specificity and sensitivity.11,12 Forexample, glucocorticoids induce the killing of thymocytes andosteoblasts but promote the survival of hepatocytes and cardiomyo-cytes. In addition, glucocorticoid sensitivity varies among subjects,within tissues of the same subject, and even within the same cellduring different stages of the cell cycle.13,14 Although cell type–specific alterations in ligand bioavailability and cofactor expressioncan modulate the glucocorticoid response, recent studies have alsoidentified a major role for an expanding array of GR isoforms.15,16

The GR is derived from a single gene; however, multiple GRproteins exist because of alternative splicing and alternativetranslation initiation mechanisms. This large cohort of function-ally distinct receptor subtypes are subject to various posttransla-tional modifications that further regulate their signalingproperties. Consequently, the cellular response to glucocorticoidsis determined in large measure by the expressed complement andcomposite actions of the individual GR isoforms. In this reviewwe discuss themolecular heterogeneity of the GR and its potentialcontribution to the regulation and dysregulation of glucocorticoidsignaling.

GLOSSARY

ACETYLATION: The addition of an acetyl group (CH3CO) to an organiccompound.

ADRENAL CORTEX: An endocrine organ made up of specific zones thatsynthesizes 3 different classes of steroids: (1) glucocorticoids (cortisol),which are synthesized primarily in the fasciculata, the middle layer; (2)

mineralocorticoids (aldosterone), which are generated in the glomer-ulosa, the outermost layer; and (3) sex steroids (estrogens and andro-gens), which are produced largely in the innermost layer, the reticularis.

ALTERNATIVE SPLICING: The process by which a given gene is splicedinto more than 1 type of mRNA molecule.

ARACHIDONIC ACID: A polyunsaturated fatty acid derived from mem-brane phospholipids by the action of cytosolic phospholipase A2.

b-ARRESTINS: A family of proteins that function to transduce andterminate (desensitize) G protein–coupled receptor signals.

CAVEOLAE: Latin for ‘‘little caves’’; a membrane compartment at thesurface of most cells capable of endocytosis and exocytosis, as well ascompartmentalizing a variety of signaling activities. Caveolin-1 is themarker protein used to isolate caveolae by means of cell fractionation.

CIRCADIAN: Latin for ‘‘approximately a day’’; biological processes underdaily rhythmicity through an internal clock. Diurnal species demonstrate

a gradual increase in plasma cortisol several hours before awakening.Plasma cortisol levels reach their nadir around midnight.

The Editors wish to thank Daniel A. Searing, MD, for preparing this glossar

GR SIGNALING

Classic GR signaling pathwayThe GR is a modular protein composed of 3 major domains: an

N-terminal transactivation domain (NTD), a central DNA-binding domain (DBD), and a C-terminal ligand-binding domain(LBD; Fig 2).17 The DBD and LBD are separated by a flexible re-gion of the protein termed the hinge region. The DBD is the mostconserved domain across the nuclear receptor family and contains2 zinc fingermotifs that recognize and bind target DNA sequencescalled glucocorticoid-responsive elements (GREs). The NTDhouses a powerful transcriptional activation function (AF1) thatinteracts with coregulators and the basal transcription machineryand is the primary site for posttranslational modifications (Fig 2).The LBD, consisting of 12 a-helices and 4 b-sheets, forms a hy-drophobic pocket for binding glucocorticoids and also contains anAF2 domain that interacts with coregulators in a ligand-dependent manner.18 Two nuclear localization signals (NL1andNL2) are located at the DBD/hinge region junction andwithinthe LBD, respectively.In the absence of hormone, the GR resides predominantly in the

cytoplasm of cells as part of a large multiprotein complex thatincludes chaperone proteins (hsp90, hsp70, and p23) andimmunophilins of the FK506 family (FKBP51 and FKBP52;Fig 3).19,20 These proteins maintain the receptor in a conforma-tion that is transcriptionally inactive but favors high-affinityligand binding. Cortisol, the most abundant endogenous gluco-corticoid in human subjects, is transported in the blood predomi-nantly bound to corticosteroid-binding globulin (CBG). CBG notonly facilitates cortisol distribution but also plays a role in itsrelease to tissues. CBG-free cortisol passively diffuses acrossthe plasma membrane; however, its bioavailability within thecell is controlled by 2 enzymes working in an opposing manner.21

11b-Hydroxysteroid dehydrogenase type 2 oxidizes cortisol into

HISTONE: Proteins that are rich in the basic amino acids lysine andarginine and complexed with DNA in chromatin.

MIFEPRISTONE: A synthetic steroid that competitively binds to theglucocorticoid receptor and progesterone receptor, blocking theireffects. Clinically, it can be used to terminate a pregnancy. At high

doses, it blocks the effects of cortisol in patientswith Cushing syndrome.

NUCLEAR LOCALIZATION SIGNALS: Factors necessary for nuclearimport (eg, transcription factors, coactivators or corepressors, DNArepair enzymes, ribosomal proteins, andmRNA processing factors) that

allow molecules to pass through nuclear pore complexes throughshuttling receptors, such as importins.

POLYMORPHISM: One of 2 or more variants of a particular DNAsequence. The most common type of polymorphism involves variationat a single base pair (single nucleotide polymorphism). Polymorphisms

can also involve long stretches of DNA.

SEX-SPECIFIC DIFFERENCE: The female/male ratio among patientswithrheumatoid arthritis is 2:1 to 3:1, and that among patients with systemiclupus erythematosus is 9:1.

SIGNAL TRANSDUCER AND ACTIVATOR OF TRANSCRIPTION

FAMILY: Transcription factors that are substrates for the Jak family oftyrosine kinases. STATs are activated by Jak-dependant tyrosinephosphorylation and form dimers that then translocate into the nucleus.

y.

CRH

ACTH

Glucocorticoids

Adrenal gland

Anterior pituitary gland

Hypothalamus

Circadian Rhythm

Stress

FIG 1. Regulation of glucocorticoid hormone secretion by the

hypothalamic-pituitary-adrenal (HPA) axis. CRH, Corticotropin-releasinghormone.


VOLUME 132, NUMBER 5

OAKLEY AND CIDLOWSKI 1035

the inactive metabolite cortisone, whereas 11b-hydroxysteroiddehydrogenase type 1 converts cortisone to cortisol (Fig 3).Changes in the expression level or activity of these enzymescan contribute to cellular differences in glucocorticoid sensitivity.In fact, inhibitors of 11b-hydroxysteroid dehydrogenase type 1have been designed to limit the adverse metabolic side effectsof increased endogenous glucocorticoids.22 In contrast to cortisol,most synthetic glucocorticoids do not bind CBG and are notmetabolized by 11b-hydroxysteroid dehydrogenase type 2. Onbinding glucocorticoids, the GR undergoes a conformationalchange, resulting in the dissociation of the associated proteins.This structural rearrangement exposes the 2 nuclear localizationsignals, and the GR is rapidly translocated into the nucleusthrough nuclear pores.Once inside the nucleus, the GR binds directly to GREs and

regulates the expression of target genes (Fig 3).23,24 Theconsensus GRE sequence GGAACAnnnTGTTCT is an imperfectpalindrome that is comprised of two 6-bp half sites. The GR bindsthis element as a homodimer, with each half site occupied by 1 re-ceptor subunit. The 3-nucleotide spacing between the 2 half sitesis strictly required for the GR to dimerize on this element. TheGRE has been shown to mediate the glucocorticoid-dependent in-duction of many genes and therefore is often referred to as an acti-vating or positive GRE. However, recent genome-wide analyseshave revealed that GR occupancy of the canonical GREs canalso lead to the repression of target genes.25 These findings sug-gest a critical role for factors outside the GRE sequence, suchas epigenetic regulators and chromatin context, in determining

the polarity of the transcriptional response. A negativeglucocorticoid-responsive element (nGRE) that mediatesglucocorticoid-dependent repression of specific genes has alsobeen recently described.26 The consensus nGRE sequenceCTCC(n)0-2GGAGA is palindromic but differs from the classicGRE in sequence, in having a variable spacer that ranges from0 to 2 nucleotides, and in being occupied by 2 GR monomersthat do not homodimerize.27 These nGREs are abundantthroughout the genome; however, more work is needed to clarifythe extent this motif is used by the GR for directly repressingtarget genes and whether this element can also be used in the acti-vation of gene expression.Global GR recruitment assays indicate that only a small

fraction of GREs are actually occupied by the receptor, and thespecific sites of GR binding vary in a tissue-specific mannerbecause of differences in chromatin accessibility and exposure ofthe GRE.28 These findings suggest that the widely varying effectsof glucocorticoids on different tissues can be attributed in part tocell type–specific differences in the chromatin landscape that in-fluence which GREs are accessible for GR binding. In addition,the concentration of glucocorticoids at which the GR bindsGREs and regulates gene expression varies throughout thegenome.29 Some GREs are occupied by the GR at very low con-centrations of glucocorticoids (hypersensitive), whereas othersrequire high doses of ligand for GR binding to occur. Both chro-matin accessibility and other DNA-binding proteins appear togovern the sensitivity of specific GREs. The identification of hy-persensitive GREs suggests that low-dose glucocorticoid therapymight provide a novel treatment option for regulating specific setsof genes that avoids the deleterious side effects of high exogenousglucocorticoids.Genome-wide analyses have also found that the majority of

GR-binding sites are located outside the promoter ofglucocorticoid-responsive genes in intergenic or intragenic re-gions often far removed from the transcription start site.30 Forexample, glucocorticoid induction of b-arrestin 1 and repressionof b-arrestin 2 occurs through an intron 1 GRE and an intron 11nGRE, respectively.31 Both b-arrestin isoforms play critical rolesin the termination and transduction of G protein–coupled receptor(GPCR) signals. By altering the ratio of b-arrestin 1 and b-ar-restin 2, glucocorticoids shift the balance of G protein– and b-ar-restin–dependent signaling responses for a given GPCR.31 Thisredirection of the GPCR signaling profile might account for thesuperior clinical efficacy of the glucocorticoid/b2-adrenergic re-ceptor (b2AR) agonist combination therapies currently used fortreating asthma and chronic obstructive pulmonary disease. Theb2AR is preferentially desensitized by b-arrestin 2. Thereforethe glucocorticoid-dependent reduction in b-arrestin 2 shouldcounteract b2AR desensitization and allow agonist-boundb2AR to promote a greater or more sustained relaxation responsein airway smooth muscle cells. In addition, lower levels of b-ar-restin 2 in lung epithelial cells should impair the proinflammatoryeffects of b2AR agonists that depend on the signaling function ofb-arrestin 2. Another example of a GR-binding site located agreat distance from the transcription start site is the intragenicnGRE recently identified in exon 6 of the GR gene that mediateshomologous downregulation of GR expression, a mechanism thatmight limit therapeutic responses to glucocorticoids.32 The loca-tion of GREs and nGREs at great distances from the transcriptionstart site suggests that these response elements can loop to the pro-moter areas of target genes to regulate transcription.

Se

r-1

34

Se

r-1

41

Se

r-2

03

Se

r-2

11

Se

r-2

26

Se

r-4

04

Lys

-2

77

Lys

-2

93

Lys

-7

03

Lys

-4

19

Lys

-4

94

Lys

-4

95

APP P P P SS US P A

7778256841241

Dimerization

Nuclear Localization

Hsp90

AF1 AF2

LBDHDBDDTN

Se

r-1

13

P

FIG 2. GR domain structure and sites of posttranslational modification. Shown are the domains of the GRand regions of the receptor involved in transactivation (AF1 and AF2), dimerization, nuclear localization, andhsp90 binding. Also depicted are the amino acid residues modified by phosphorylation (P), sumoylation

(S), ubiquitination (U), and acetylation (A). Numbers are for human GR.

FIG 3. GR signaling pathways. The glucocorticoid-activated GR regulates gene expression in 3 primaryways: binding directly to DNA (A), tethering itself to other DNA-bound transcription factors (B), or binding

directly to DNA and interacting with neighboring DNA-bound transcription factors (C). The GR can alsosignal in a nongenomic manner through alterations in the activity of various kinases. BTM, Basal transcrip-tion machinery; PI3K, phosphoinositide 3-kinase; STAT, signal transducer and activator of transcription.


NOVEMBER 2013


The interaction of the GR with DNA is highly dynamic, withthe GR cycling between bound and unbound states every fewseconds.33 Once bound to the GRE, the receptor undergoes addi-tional conformational changes that lead to the recruitment of cor-egulators and chromatin-remodeling complexes that modulate

gene transcription rates by affecting the activity of RNA polymer-ase II.34-36 Cofactors that mediate transcriptional activationinclude steroid receptor coactivators, the histone acetyltransfer-ase CBP/p300, and the nuclear methylase coactivator-associatedarginine methyltransferase. Nuclear receptor corepressor 1 and




silencing mediator of retinoid and thyroid hormone receptors areestablished corepressors that are recruited to the GR bound tonGREs. The specific cofactors assembled and their subsequentactivity are dictated by both the nature of the glucocorticoidligand and the specific GRE sequence bound by the receptor.37,38

The GR can also regulate the transcription of target genes byphysically interacting with other transcription factors (Fig 3). Theassociation of the GR with specific members of the signal trans-ducer and activator of transcription family, either apart from orin conjunction with GRE binding, has been shown to enhance thetranscription of responsive genes.39 In contrast, the interaction ofthe GR with the proinflammatory transcription factors activatorprotein 1 (AP1) and nuclear factor kB (NF-kB) antagonizes theiractivity and is considered to be a primary mechanism throughwhich glucocorticoids suppress inflammation. The GR directlybinds the Jun subunit of AP1 and the p65 subunit of NF-kB andinterferes with the transcriptional activation function of these 2proteins.40,41 For some genes, the repression is accomplished bythe GR tethering itself to these DNA-bound proteins without itselfdirectly interacting with the DNA. For other genes, however, theGR functions in a composite manner, binding directly to a GREand physically associating with AP1 or NF-kB bound to a neigh-boring site on the DNA. GR-dependent recruitment of the gluco-corticoid receptor–interacting protein 1, a transcriptionalcoregulator of the p160 family, is important for this inhibition.42

Interestingly, recent work has demonstrated that the anti-inflammatory actions of glucocorticoids differ in a sex-specificmanner.43 Glucocorticoids regulate a greater number of inflam-matory genes and elicit a greater anti-inflammatory response inmale compared with female rats. Sex-specific expression profilesof coregulatory molecules might underlie the sexual dimorphismby modulating the repressive actions of the GR on AP1 and NF-kB. The sex-specific differences in the anti-inflammatory actionsof glucocorticoids not only provide a mechanistic basis for thepredisposition of female subjects to autoimmune diseases, suchas rheumatoid arthritis and systemic lupus erythematosus, butalso suggest that therapeutic doses of glucocorticoids should beoptimized according to the sex of the patient.

Nonclassical GR signaling pathwayAlthough the principal effects of glucocorticoids are mediated

by transcriptional responses that occur in minutes to hours, agrowing body of evidence suggests that the GR can also actthrough nongenomic mechanisms to elicit rapid cellular re-sponses that occur within a few seconds to minutes and do notrequire changes in gene expression (Fig 3).44,45 Multiple mecha-nisms appear to be involved in these signaling events that ulti-mately impinge on the activity of various kinases, such asphosphoinositide 3-kinase, AKT, and mitogen-activated proteinkinases (MAPKs). For example, the glucocorticoid-dependentrelease of accessory proteins associated with unliganded GR inthe cytoplasm, such as the nonreceptor tyrosine kinase c-Src,can mediate nongenomic actions of glucocorticoids. Whenliberated from the GR complex, c-Src activates multiple kinasecascades that lead to the phosphorylation of annexin 1, inhibitionof cytosolic phospholipase A2 activity, and impaired releaseof arachidonic acid.46,47 The GR has also been reported tolocalize at the plasma membrane in caveolae through an interac-tion with caveolin 1.48 Glucocorticoid activation of thismembrane-associated GR regulates gap junction intercellular

communication and neural progenitor cell proliferation througha mechanism that requires c-Src activity and rapid MAPK-dependent phosphorylation of connexin-43.49 The existence ofnongenomic signaling adds greater complexity and diversity toglucocorticoids and their biological actions and raises the possi-bility that selectivemodulators of GR-dependent genomic or non-genomic pathways might be therapeutically advantageous.

GR polymorphismsThe capacity of the GR to function as a transcriptional activator

or repressor is affected by several polymorphisms in the GR genethat alter the amino acid sequence of the encoded receptor. TheER22/23EK GR polymorphism occurs in approximately 3% ofthe population and results in an arginine to lysine change at posi-tion 23 within the NTD.50,51 GR with the ER22/23EK polymor-phism exhibits reduced transcriptional activity on bothglucocorticoid-responsive reporters and endogenous genes andhas been associated with glucocorticoid insensitivity. Personswith the ER22/23EK polymorphism have a lower tendency tohave impaired glucose tolerance, type 2 diabetes, and cardiovas-cular disease, suggesting the relative resistance of ER22/23EKcarriers to endogenous glucocorticoids might result in a morefavorable metabolic profile.The N363S polymorphism occurs in approximately 4% of the

population and results in an asparagine to serine substitution intheNTDof theGR.50 In contrast to ER22/23EK, the presence of theN363S polymorphism enhances the transcriptional activity of theGR and is associated with glucocorticoid hypersensitivity. Inaddition, a genome-widemicroarray analysis has revealed a uniquepattern of gene regulation for the N363S polymorphism.52 N363Scarriers have been reported to have an increased body mass indexand a tendency toward decreased bone mineral density.51

These GR polymorphisms can account, at least in part, for thevariability in the glucocorticoid response observed amongsubjects treated with these steroids.

GR SPLICE VARIANTSThe human GR gene is composed of 9 exons. The GR NTD is

encoded primarily by exon 2, the DBD is encoded by exons 3 and4, and the hinge region and LBD are encoded by exons 5 to 9(Fig 4). Alternative splicing in exon 9 near the end of the GR pri-mary transcript generates 2 receptor isoforms, termed GRa andGRb (Fig 4).53,54 GRa is derived from the end of exon 8 beingjoined to the beginning of exon 9, whereas GRb results fromthe end of exon 8 being joined to downstream sequences inexon 9. These 2 proteins are identical through amino acid 727but then diverge. GRa, the classic GR protein that mediates theactions of glucocorticoids, contains an additional 50 amino acids.The splice variant GRb contains an additional, nonhomologous15 amino acids. The unique GRb carboxyl-terminal sequenceconfers several distinct properties to this receptor isoform. GRbdoes not bind glucocorticoid agonists, resides constitutively inthe nucleus of cells, and by itself is inactive on glucocorticoid-responsive reporter genes.55,56 However, when coexpressedwith GRa, GRb functions as a dominant negative inhibitor andantagonizes the activity of GRa on many glucocorticoid-responsive target genes. Competition for GRE binding, competi-tion for transcriptional coregulators, and formation of inactiveGRa/GRb heterodimers have each been proposed to underliethe antagonism mediated by GRb.

FIG 4. GR splice variants. The GR primary transcript is composed of 9 exons. Exon 2 encodes the NTD,exons 3 to 4 encode the DBD, and exons 5 to 9 encode the hinge region (H) and LBD. GRa results fromsplicing exon 8 to the beginning of exon 9. GRb, GRg, GR-A, and GR-P are generated by the depictedalternative splicing events.


NOVEMBER 2013


The ability of GRb to inhibit the activity of GRa suggests thathigh levels of GRb might lead to glucocorticoid resistance. GRbis prevalent in many cells and tissues but generally is found atlower levels than GRa. However, GRb is abundant in certain celltypes, such as neutrophils and epithelial cells. Moreover, theexpression of GRb can be selectively increased by proinflamma-tory cytokines and other immune activators and lead to reducedglucocorticoid sensitivity.57-59 Increased GRb levels have beenassociated with glucocorticoid resistance in a variety of inflam-matory diseases, including asthma, rheumatoid arthritis, ulcera-tive colitis, nasal polyposis, systemic lupus erythematosus,sepsis, acute lymphoblastic leukemia, and chronic lymphocyticleukemia.56 Upregulation of GRb and diminished GRa signalingare also observed in erythroid cells expanded from patients withpolycythemia vera.60 The molecular factors that control GRbexpression are poorly understood, but several studies have impli-cated the involvement of the splicing factor SRp30c.61-63 Treat-ment of cells with agents that induce SRp30c expression leadsto a selective increase in GRb levels and consequent glucocorti-coid insensitivity. A naturally occurring polymorphism(A3669G) in the 39 untranslated region of the GRb mRNA alsoleads to increased GRb expression.64,65 The polymorphism dis-rupts an mRNA destabilization motif (AUUUA), resulting in aprolonged half-life of the GRb mRNA. A3669G carriers havean increased risk for pathologies with known inflammatory under-pinnings, such as autoimmune disease, myocardial infarction,coronary artery disease, and heart failure.66-68 These findings sug-gest that the increase in GRb levels in persons with the A3669Gpolymorphismmight attenuate the beneficial immunosuppressiveand anti-inflammatory actions of GRa.Recent discoveries suggest an even broader role for GRb in cell

signaling and physiology. Using genome-wide microarrays oncells engineered to selectively express GRb, multiple laboratorieshave shown that GRb directly induces and represses theexpression of a large number of genes independent of its dominant

negative activity on GRa.69,70 The ability of GRb to recruit his-tone deacetylases and close local chromatin structures appearsto be involved in its repression of certain genes, such as IL-5and IL-13.71,72 GRb can also bind the glucocorticoid antagonistmifepristone (RU486).70 In the presence of RU486, many ofthe GRb-mediated changes in gene expression are abolished.These findings indicate that GRb can function as a bona fide tran-scription factor, and they raise the possibility that GRbmodulatesglucocorticoid responses through genomic actions distinct fromits antagonism of GRa. Indeed, GRb has recently been shownto directly reduce the expression of histone deacetylase 2 and pro-mote glucocorticoid insensitivity in bronchoalveolar lavage cellsfrom patients with steroid-resistant asthma.73

Another interesting finding has been the discovery of GRb inzebrafish, mice, and rats.74-76 These isoforms are similar in struc-ture and function to human GRb but originate from a mechanismof alternative splicing that differs from that in human subjects.Knockout of GRb in these species promises to shed new lighton the physiologic and pathologic importance of this splicevariant.In addition to GRb, alternative splicing of the GR gene gives

rise to other receptor isoforms with distinct signaling properties(Fig 4). The use of an alternative splice donor site in the intronseparating exons 3 and 4 results in a receptor isoform that containsan insertion of a single arginine residue between the 2 zinc fingersof the GR DBD.77 This receptor variant, termed GRg, is widelyexpressed and binds glucocorticoids andDNA in amanner similarto GRa. However, GRg is impaired in its ability to regulateglucocorticoid-responsive reporter genes and exhibits a transcrip-tional profile distinct from GRa on a subset of endogenous genesregulated by glucocorticoids. GRg expression is associatedwith glucocorticoid resistance in patients with small celllung carcinoma, corticotroph adenomas, and childhood acutelymphoblastic leukemia.77-79 Two GR splice variants missinglarge regions of the LBD were initially discovered in

FIG 5. GR translational isoforms. Translation initiation from 8 different AUG start codons in the single GRamRNA (location of the AUG start codons indicated by an asterisk) produces 8 receptor isoforms with pro-gressively shorter amino NTDs. The numbers for the initiator methionines and AF1 region (amino acids77-262) are for the human GRa protein. UTR, Untranslated region.




glucocorticoid-resistant multiple myeloma cells.80 GR-A ismissing middle exons 5 to 7, which encode the amino-terminalhalf of the LBD, and GR-P is missing the terminal exons 8 to 9,which encode the carboxyl-terminal half of the LBD. As expectedfrom these changes in the LBD, GR-A and GR-P do not bind glu-cocorticoids. Little is currently known about GR-A; however,GR-P has been shown to modulate the transcriptional activity ofGRa in a cell type–specific manner.81-83 GR-P is found in manytissues and appears to be the predominant receptor variantexpressed in glucocorticoid-resistant cancer cells.

GR TRANSLATIONAL ISOFORMSAlternative translation initiation from the single GRa mRNA

transcript produces an additional cohort of diverse GR proteins(Fig 5). Eight AUG start codons derived from exon 2 of the GRgene give rise to 8 GRa subtypes with progressively shorterNTDs (GRa-A, GRa-B, GRa-C1, GRa-C2, GRa-C3,GRa-D1, GRa-D2, and GRa-D3).15,84 All 8 sites are highlyconserved across species. GRa-A is the full-length receptor thatis generated from the first initiator codon. Production of the trun-cated isoforms from internal start codons involves both ribosomalleaky scanning and ribosomal shunting mechanisms. The mRNAspecies encoding the splice variants discussed above (GRb, GRg,GR-A, andGR-P) also contain the identical set of start codons andwould therefore be expected to give rise to a similar complementof translational isoforms.

The GRa translational isoforms display a similar affinity forglucocorticoids and a similar capacity to interact with GREs afterligand activation; however, they showmarked differences in otherproperties.8,15 The GRa-A, GRa-B, and GRa-C isoforms arelocalized in the cytoplasm of cells in the absence of hormoneand translocate to the nucleus on glucocorticoid binding. Incontrast, the GRa-D isoforms reside constitutively in the nucleiof cells. In addition, nuclear localized GRa-D associates withGRE-containing promoters of specific target genes independentof glucocorticoid treatment. Most striking, however, is that eachGRa subtype possesses a distinct gene regulatory profile. Whenthe individual isoforms are expressed at similar levels in U2OSosteosarcoma or Jurkat T lymphoblastic leukemia cells, theyeach regulate a unique set of genes.8,85 Less than 10%of the geneswere commonly regulated by all the subtypes, indicating that thevast majority of genes were selectively regulated by differentGRa isoforms. These isoform-unique gene regulatory profileslead to functional differences in glucocorticoid-inducedapoptosis.8,85,86 Cells expressing GRa-C3 were the most sensi-tive to the cell-killing effects of glucocorticoids, whereas cells ex-pressing GRa-D3 were the most resistant. The heightenedactivity of the GRa-C3 isoform has also been observed onglucocorticoid-responsive reporters and has been linked to anN-terminal motif (residues 98-115) that is sterically hindered inthe larger receptor isoforms.15,87 When unobstructed inGRa-C3, this motif increases the activity of the AF1 domain, pre-sumably through enhanced recruitment of various coregulators.


NOVEMBER 2013


The complete absence of the AF-1 domain likely underlies thereduced transcriptional activity of the GRa-D isoforms. GRa-D3 does not efficiently repress the expression of multiple antia-poptotic genes because of deficits in its ability to interact withthe p65 subunit of NF-kB and to be recruited to the promotersof these glucocorticoid-responsive genes.86

The discovery that each GRa translational isoform regulates aunique transcriptome suggests that the cellular response toglucocorticoids will be governed by the expressed complementof receptor subtypes. The GRa-A and GRa-B isoforms are themost abundant GR proteins in many cell types. However,trabecular meshwork cells from the human eye preferentiallyexpress the GRa-C and GRa-D isoforms.88 In addition, immaturedendritic cells predominantly express the GRa-D isoforms,whereas mature dendritic cells predominantly express the GRa-Asubtype.89 In rodents levels of GRa-C isoforms are highest in thepancreas and colon, and levels of the GRa-D subtypes are highestin the spleen and lung.15Recent studies have also demonstrated thatthe composition of GRa subtypes in a given cell or tissue canchange in response to different conditions. Treatment of differenti-ated murine skeletal muscle cells with a selective estrogen-relatedreceptor b/g agonist induced a preferential increase in levels of theGRa-D isoforms,90 and levels of the GRa-C isoforms are selec-tively upregulated after mitogen activation of human primary Tcells.85 In the human brain alterations in the expression of the GRsubtypes were observed during development and during the ageingprocess.91 Moreover, a selective increase in GRa-D levels wasmeasured in specific regions of the brain in patients with schizo-phrenia and bipolar disorder.92,93 Because preferential increasesin the expression of the GRa-C or GRa-D isoforms might lead toglucocorticoid hypersensitivity or hyposensitivity, respectively,an important goal of future studies will be to define the operativecellular mechanisms that control the expressed complement oftranslational isoforms.Currently, the efficiencywithwhich a partic-ular start codon is used has been reported to be influenced by theER22/23EK polymorphism in the GR gene and heterogeneity inthe 59 untranslated region of theGRamRNA.94,95 Posttranslationalmodifications that differentially affect the half-life of thevarious re-ceptor subtypes might also contribute to alterations in the expres-sion profile of the GRa isoforms.

POSTTRANSLATIONAL MODIFICATION OF GR

ISOFORMSEach GR isoform originating from alternative processing of the

GR gene is subject to a variety of posttranslational modificationsthat further modulate its function and expand the repertoire ofreceptor subtypes available for glucocorticoid signaling. The firstidentified and most extensively studied covalent modification ofGR is phosphorylation.96-98 Human GRa is phosphorylated on atleast 7 serine residues (Ser-113, Ser-134, Ser-141, Ser-203, Ser-211, Ser-226, and Ser-404), all of which are located in the NTDof the receptor (Fig 2). The major kinases that phosphorylateGRa include MAPKs, cyclin-dependent kinases, casein kinaseII, and glycogen synthase kinase 3b. Many of the sites exhibit alow level of basal phosphorylation and become hyperphosphory-lated after the binding of glucocorticoids.99 The structure of thebound glucocorticoid can influence both the pattern and extentof GRa phosphorylation.100 Ser-134 is unique in that it is phos-phorylated in a glucocorticoid-independent manner by stress-activating stimuli.101

One of the major effects of GRa phosphorylation is that itchanges the transcriptional activity of the receptor. Early studiesdemonstrated that phosphorylation-deficient GRa mutants wereimpaired in their ability to activate some glucocorticoid-responsive promoters but not others.102 Subsequent reportsshowed that phosphorylation at Ser-211 correlated with increasedtranscriptional activity of GRa, whereas phosphorylation at Ser-226 decreased the signaling capacity of the receptor.99,103,104

A deficiency in Ser-211 phosphorylation might contribute to theresistance to glucocorticoid-induced apoptosis that develops inmalignant lymphoid cells.105,106 Conversely, increased phosphor-ylation at Ser-226 might account for the impaired glucocorticoidsignaling in the pathophysiology of depression.107

Phosphorylation of Ser-404 also has major consequences onglucocorticoid responses because the ability of GRa to bothactivate and repress target genes is diminished by phosphoryla-tion at this site.108 In cells expressing a GRamutant incapable ofSer-404 phosphorylation, the global transcriptional response toglucocorticoids is redirected to favor the activation of distinctsignaling pathways.Consistent with many of the phosphorylation sites residing

within the AF1 domain, differences in cofactor recruitmentappear to be responsible for the transcriptional alterations thataccompany GRa phosphorylation. GRa recruitment of thecoactivator MED14 is enhanced by glucocorticoid-dependentphosphorylation at Ser-211,104 whereas the interaction of GRawith the coactivator p300/CBP and the p65 subunit of NF-kBare both diminished by phosphorylation at Ser-404.108

Phosphorylation alters other properties of GR that effect theprofile of glucocorticoid signaling. Degradation of the GRaprotein is enhanced by glucocorticoid-dependent phosphoryla-tion of the receptor because phosphorylation-deficient mutantsare stabilized in the presence of glucocorticoids.102 In addition,the ligand-free GRa is protected from degradation by its associ-ation with the tumor-suppressor gene TSG101, which preferen-tially interacts with the nonphosphorylated receptor.109 Thecellular distribution of GRa is also altered by receptor phosphor-ylation. GRa phosphorylated on Ser-203, Ser-226, or Ser-404spends less time in the nuclear compartment because of greatercytoplasmic retention, enhanced nucleocytoplasmic transport,or both. As a consequence, GRa phosphorylated on either of theseresidues exhibits reduced transcriptional activity onglucocorticoid-responsive target genes.99,103,108,110

Additional posttranslational modifications of the GR have beendescribed that regulate the function of the receptor. Ubiquitin is a76-amino-acid protein that, when attached to specific lysineresidues, marks proteins for proteasomal degradation. Ubiquiti-nation of GRa at a conserved lysine residue (Lys-419) has beenshown to target the receptor for turnover by the proteasome(Fig 2).111,112Mutant receptors that cannot be ubiquitinated at thisresidue are resistant to ligand-dependent downregulation andexhibit potentiated transcriptional activity on glucocorticoid-responsive reporter genes. Moreover, changes in the expressionof an E3 ubiquitin ligase for GRa leads to alterations in bothreceptor levels and cellular sensitivity to glucocorticoids.113

GRa is also posttranslationally modified by sumoylation, aprocess in which SUMO (small ubiquitin-related modifier)peptides are covalently attached to specific lysine residues (Lys-277, Lys-293, Lys-703) within the receptor (Fig 2). The additionof SUMO peptides to the receptor occurs in the absence of ligandand is increased by the binding of glucocorticoids. Depending on




the site of sumoylation, the transcriptional activity of GRa can beenhanced or repressed through alterations in the recruitment oractivity of specific coregulators.114-119

Finally, it has been suggested that GRa becomes acetylated onlysine residues (Lys-494 and Lys-495) located within the hingeregion in response to glucocorticoid binding (Fig 2).120 Deacety-lation of the GR by histone deacetylase 2 was reported to berequired for the receptor to efficiently repress the transcriptionalactivity of NF-kB, suggesting that acetylation of the GR limitsthe inhibitory actions of glucocorticoids on NF-kB signaling.More recently, studies have shown that the clock transcriptionfactor acetylates GRa and attenuates its ability to both activateand repress glucocorticoid responsive genes, conferring glucocor-ticoid insensitivity to target tissues.121

Clearly, posttranslational modifications can regulate multipleaspects of GRa function, providing cells with additional receptorheterogeneity for controlling glucocorticoid responses. Animportant goal of future research will be to determine to whatextent the various splicing and translational isoforms of the GRare subject to and regulated by these posttranslationalmodifications.

SUMMARY AND FUTURE PERSPECTIVESGlucocorticoids act through the GR to regulate numerous

physiologic processes, and synthetic derivatives of these hor-mones are widely prescribed for treating inflammatory diseases,autoimmune disorders, and various cancers. A challenge toclinicians and patients alike is that cell, tissue, and individualresponses to glucocorticoids are markedly different and canchange over time and in response to various extracellular cues.The discovery that multiple GR isoforms arise from a single genehas advanced our understanding of the molecular basis for thediversity in glucocorticoid signaling. GR isoforms with uniqueexpression, gene-regulatory, and functional profiles are generatedby alternative splicing of the primary transcript, alternativetranslation initiation of the mature mRNA, and posttranslationalmodifications of the encoded protein. The potential for these GRsubtypes to undergo various posttranslational modifications andto function as monomers, homodimers, and heterodimers pro-vides cells with a wealth of possibilities for generating diverseglucocorticoid responses with fine-tuned precision. Alterationsin the relative levels of the GR subtypes might underliepathologies characterized by hyposensitivity or hypersensitivityto glucocorticoids.The cellular response to glucocorticoids will depend not only

on the GR isoform composition but also on the glucocorticoid thatbinds and activates the receptor, as well as the concentration of theadministered steroid. Not all glucocorticoids are created equalbecause structurally different but similarly potent steroids used inthe clinic regulate both common and unique sets of genes.88,122

The distinct transcriptional signature of these glucocorticoidssuggests that their binding confers unique conformations on theGR that lead to differences in DNA binding, chromatin remodel-ing, and/or coregulator recruitment. Recent discoveries also sug-gest that the concentration of glucocorticoids needed to achievethe desired transcriptional response can vary in a context-specific manner. For example, GR repression of proinflammatorygenes through tethering to NF-kB occurs at much lower glucocor-ticoid concentrations than the induction of gene expression.123

Global GR recruitment assays have revealed that different

GREs and their associated target genes also exhibit profound dif-ferences in their sensitivity to glucocorticoids.29 In addition,dose-response studies in the livers of male and female rats havediscovered several genes in female rats that are 10- to 100-foldless sensitive to glucocorticoid regulation than they are in malerats.43 Remarkably, even the polarity of the transcriptionalresponse of certain GR-regulated genes can vary depending onthe dose of glucocorticoid.124 Collectively, these findings suggestthat as the specific genes underlying various pathologies becomeidentified, careful consideration of the structure and dose of theused glucocorticoid might help fine-tune the glucocorticoidresponse and lead to improved benefit/risk ratios for patients un-dergoing steroid therapy.Glucocorticoid resistance remains a major barrier to the

effective treatment of a variety of immune and inflammatorydiseases.125,126 Advances in our understanding of glucocorticoidsignaling have uncovered a variety of mechanisms that contributeto reduced glucocorticoid responsiveness, including increasedexpression of the GRb andGRa-D isoforms, changes inGRphos-phorylation, and homologous downregulation of the receptor. Un-derstanding the molecular mechanisms of resistance permits notonly the prediction of patient responsiveness to glucocorticoidsbut also the design of novel therapeutic strategies for combatingthe insensitivity. For instance, altered patterns of GR phosphory-lation thought to contribute to glucocorticoid resistance in pa-tients with severe asthma can be reversed by coadministrationof long-acting b2AR agonists or p38MAPK inhibitors.125 Devel-opment of glucocorticoid analogues deficient in their ability todownregulate the expression of GR might also prove to be aneffective strategy for overcoming steroid resistance.The harmful side effects of medicinal glucocorticoids also

remain a frustration for clinicians and their patients. Thereforeintense efforts have been made over the last decade to developnovel GR ligands, termed dissociated or selective glucocorticoidreceptor agonists (SEGRAs), that retain the negative regulation ofgene expression thought to be important for many of the anti-inflammatory actions of glucocorticoids but lose the positiveregulation of gene expression thought to underlie many of theiradverse effects. A number of SEGRAs have been developed, butfew have made it to clinical trials, and several recent studies havechallenged both the premise and utility of these analogues astherapeutic agents.127,128 One potential shortcoming of SEGRAsis that mounting evidence indicates that glucocorticoid-dependent activation of many genes also plays a critical role inthe anti-inflammatory mechanisms of glucocorticoids. Clearly,more research is needed to unravel the molecular details by whichglucocorticoids repress the inflammatory response and induce un-desirable side effects. A greater understanding of the heterogene-ity in GR signaling responses in both healthy and diseased tissueswill aid in the development of safer and more effective glucocor-ticoid therapies.

REFERENCES

1. Barnes PJ. Anti-inflammatory actions of glucocorticoids: molecular mechanisms.

Clin Sci (Lond) 1998;94:557-72.

2. Sapolsky RM, Romero LM, Munck AU. How do glucocorticoids influence stress

responses? Integrating permissive, suppressive, stimulatory, and preparative

actions. Endocr Rev 2000;21:55-89.

3. Rhen T, Cidlowski JA. Antiinflammatory action of glucocorticoids—new mech-

anisms for old drugs. N Engl J Med 2005;353:1711-23.

4. Busillo JM, Cidlowski JA. The five Rs of glucocorticoid action during inflamma-

tion: ready, reinforce, repress, resolve, and restore. Trends Endocrinol Metab

2013;24:109-19.

http://refhub.elsevier.com/S0091-6749(13)01388-2/sref1










christinehurd

Highlight

christinehurd

Highlight


NOVEMBER 2013


5. Miner JN, Hong MH, Negro-Vilar A. New and improved glucocorticoid receptor

ligands. Expert Opin Investig Drugs 2005;14:1527-45.

6. Evans RM. The steroid and thyroid hormone receptor superfamily. Science 1988;

240:889-95.

7. Galon J, Franchimont D, Hiroi N, Frey G, Boettner A, Ehrhart-Bornstein M, et al.

Gene profiling reveals unknown enhancing and suppressive actions of glucocor-

ticoids on immune cells. FASEB J 2002;16:61-71.

8. Lu NZ, Collins JB, Grissom SF, Cidlowski JA. Selective regulation of bone cell

apoptosis by translational isoforms of the glucocorticoid receptor. Mol Cell Biol

2007;27:7143-60.

9. Ren R, Oakley RH, Cruz-Topete D, Cidlowski JA. Dual role for glucocorti-

coids in cardiomyocyte hypertrophy and apoptosis. Endocrinology 2012;153:

5346-60.

10. Cole TJ, Blendy JA, Monaghan AP, Krieglstein K, Schmid W, Aguzzi A, et al.

Targeted disruption of the glucocorticoid receptor gene blocks adrenergic chro-

maffin cell development and severely retards lung maturation. Genes Dev

1995;9:1608-21.

11. Kino T, De Martino MU, Charmandari E, Mirani M, Chrousos GP. Tissue gluco-

corticoid resistance/hypersensitivity syndromes. J Steroid Biochem Mol Biol

2003;85:457-67.

12. Lamberts SW, Huizenga AT, de Lange P, de Jong FH, Koper JW. Clinical aspects

of glucocorticoid sensitivity. Steroids 1996;61:157-60.

13. Gorovits R, Ben-Dror I, Fox LE, Westphal HM, Vardimon L. Developmental

changes in the expression and compartmentalization of the glucocorticoid recep-

tor in embryonic retina. Proc Natl Acad Sci U S A 1994;91:4786-90.

14. Hsu SC, DeFranco DB. Selectivity of cell cycle regulation of glucocorticoid re-

ceptor function. J Biol Chem 1995;270:3359-64.

15. Lu NZ, Cidlowski JA. Translational regulatory mechanisms generate N-terminal

glucocorticoid receptor isoforms with unique transcriptional target genes. Mol

Cell 2005;18:331-42.

16. Oakley RH, Cidlowski JA. Cellular processing of the glucocorticoid receptor gene

and protein: new mechanisms for generating tissue-specific actions of glucocorti-

coids. J Biol Chem 2011;286:3177-84.

17. Kumar R, Thompson EB. Gene regulation by the glucocorticoid receptor: struc-

ture:function relationship. J Steroid Biochem Mol Biol 2005;94:383-94.

18. Bledsoe RK, Montana VG, Stanley TB, Delves CJ, Apolito CJ, McKee DD, et al.

Crystal structure of the glucocorticoid receptor ligand binding domain reveals a

novel mode of receptor dimerization and coactivator recognition. Cell 2002;

110:93-105.

19. Grad I, Picard D. The glucocorticoid responses are shaped by molecular chaper-

ones. Mol Cell Endocrinol 2007;275:2-12.

20. Pratt WB, Toft DO. Steroid receptor interactions with heat shock protein and im-

munophilin chaperones. Endocr Rev 1997;18:306-60.

21. Seckl JR. 11beta-hydroxysteroid dehydrogenases: changing glucocorticoid ac-

tion. Curr Opin Pharmacol 2004;4:597-602.

22. Yang H, Dou W, Lou J, Leng Y, Shen J. Discovery of novel inhibitors of 11beta-

hydroxysteroid dehydrogenase type 1 by docking and pharmacophore modeling.

Bioorg Med Chem Lett 2008;18:1340-5.

23. Beato M. Gene regulation by steroid hormones. Cell 1989;56:335-44.

24. Freedman LP. Anatomy of the steroid receptor zinc finger region. Endocr Rev

1992;13:129-45.

25. Uhlenhaut NH, Barish GD, Yu RT, Downes M, Karunasiri M, Liddle C, et al. In-

sights into negative regulation by the glucocorticoid receptor from genome-wide

profiling of inflammatory cistromes. Mol Cell 2013;49:158-71.

26. Surjit M, Ganti KP, Mukherji A, Ye T, Hua G, Metzger D, et al. Widespread nega-

tive response elements mediate direct repression by agonist-liganded glucocorti-

coid receptor. Cell 2011;145:224-41.

27. Hudson WH, Youn C, Ortlund EA. The structural basis of direct glucocorticoid-

mediated transrepression. Nat Struct Mol Biol 2013;20:53-8.

28. John S, Sabo PJ, Thurman RE, Sung MH, Biddie SC, Johnson TA, et al. Chro-

matin accessibility pre-determines glucocorticoid receptor binding patterns. Nat

Genet 2011;43:264-8.

29. Reddy TE, Gertz J, Crawford GE, Garabedian MJ, Myers RM. The hypersensitive

glucocorticoid response specifically regulates period 1 and expression of circa-

dian genes. Mol Cell Biol 2012;32:3756-67.

30. Burd CJ, Archer TK. Chromatin architecture defines the glucocorticoid response.

Mol Cell Endocrinol 2013 [Epub ahead of print].

31. Oakley RH, Revollo J, Cidlowski JA. Glucocorticoids regulate arrestin gene

expression and redirect the signaling profile of G protein-coupled receptors.

Proc Natl Acad Sci U S A 2012;109:17591-6.

32. Ramamoorthy S, Cidlowski JA. Ligand-induced repression of the glucocorticoid

receptor gene is mediated by an NCoR1 repression complex formed by long-

range chromatin interactions with intragenic glucocorticoid response elements.

Mol Cell Biol 2013;33:1711-22.

33. McNally JG, Muller WG, Walker D, Wolford R, Hager GL. The glucocorticoid

receptor: rapid exchange with regulatory sites in living cells. Science 2000;287:

1262-5.

34. Jenkins BD, Pullen CB, Darimont BD. Novel glucocorticoid receptor coactivator

effector mechanisms. Trends Endocrinol Metab 2001;12:122-6.

35. Lonard DM, O’Malley BW. Expanding functional diversity of the coactivators.

Trends Biochem Sci 2005;30:126-32.

36. Rosenfeld MG, Glass CK. Coregulator codes of transcriptional regulation by nu-

clear receptors. J Biol Chem 2001;276:36865-8.

37. Meijsing SH, Pufall MA, So AY, Bates DL, Chen L, Yamamoto KR. DNA bind-

ing site sequence directs glucocorticoid receptor structure and activity. Science

2009;324:407-10.

38. Ronacher K, Hadley K, Avenant C, Stubsrud E, Simons SS Jr, Louw A, et al.

Ligand-selective transactivation and transrepression via the glucocorticoid recep-

tor: role of cofactor interaction. Mol Cell Endocrinol 2009;299:219-31.

39. Rogatsky I, Ivashkiv LB. Glucocorticoid modulation of cytokine signaling. Tissue

Antigens 2006;68:1-12.

40. Nissen RM, Yamamoto KR. The glucocorticoid receptor inhibits NFkappaB by

interfering with serine-2 phosphorylation of the RNA polymerase II carboxy-

terminal domain. Genes Dev 2000;14:2314-29.

41. Yang-Yen HF, Chambard JC, Sun YL, Smeal T, Schmidt TJ, Drouin J, et al. Tran-

scriptional interference between c-Jun and the glucocorticoid receptor: mutual in-

hibition of DNA binding due to direct protein-protein interaction. Cell 1990;62:

1205-15.

42. Chinenov Y, Gupte R, Dobrovolna J, Flammer JR, Liu B, Michelassi FE, et al.

Role of transcriptional coregulator GRIP1 in the anti-inflammatory actions of glu-

cocorticoids. Proc Natl Acad Sci U S A 2012;109:11776-81.

43. Duma D, Collins JB, Chou JW, Cidlowski JA. Sexually dimorphic actions of glu-

cocorticoids provide a link to inflammatory diseases with gender differences in

prevalence. Sci Signal 2010;3:ra74.

44. Groeneweg FL, Karst H, de Kloet ER, Joels M. Mineralocorticoid and glucocor-

ticoid receptors at the neuronal membrane, regulators of nongenomic corticoste-

roid signalling. Mol Cell Endocrinol 2012;350:299-309.

45. Samarasinghe RA, Witchell SF, DeFranco DB. Cooperativity and complemen-

tarity: synergies in non-classical and classical glucocorticoid signaling. Cell Cy-

cle 2012;11:2819-27.

46. Croxtall JD, Choudhury Q, Flower RJ. Glucocorticoids act within minutes to

inhibit recruitment of signalling factors to activated EGF receptors through a

receptor-dependent, transcription-independent mechanism. Br J Pharmacol

2000;130:289-98.

47. Solito E,MullaA,Morris JF, Christian HC, Flower RJ, Buckingham JC. Dexameth-

asone induces rapid serine-phosphorylation and membrane translocation of an-

nexin 1 in a human folliculostellate cell line via a novel nongenomic mechanism

involving the glucocorticoid receptor, protein kinase C, phosphatidylinositol

3-kinase, and mitogen-activated protein kinase. Endocrinology 2003;144:1164-74.

48. Matthews L, Berry A, Ohanian V, Ohanian J, Garside H, Ray D. Caveolin medi-

ates rapid glucocorticoid effects and couples glucocorticoid action to the antipro-

liferative program. Mol Endocrinol 2008;22:1320-30.

49. Samarasinghe RA, Di Maio R, Volonte D, Galbiati F, Lewis M, Romero G, et al.

Nongenomic glucocorticoid receptor action regulates gap junction intercellular

communication and neural progenitor cell proliferation. Proc Natl Acad Sci U

S A 2011;108:16657-62.

50. Gross KL, Cidlowski JA. Tissue-specific glucocorticoid action: a family affair.

Trends Endocrinol Metab 2008;19(9):331-9.

51. van Rossum EF, Lamberts SW. Polymorphisms in the glucocorticoid receptor

gene and their associations with metabolic parameters and body composition.

Recent Prog Horm Res 2004;59:333-57.

52. Jewell CM, Cidlowski JA. Molecular evidence for a link between the N363S

glucocorticoid receptor polymorphism and altered gene expression. J Clin Endo-

crinol Metab 2007;92:3268-77.

53. Bamberger CM, Bamberger AM, de Castro M, Chrousos GP. Glucocorticoid re-

ceptor beta, a potential endogenous inhibitor of glucocorticoid action in humans.

J Clin Invest 1995;95:2435-41.

54. Oakley RH, Sar M, Cidlowski JA. The human glucocorticoid receptor beta iso-

form. Expression, biochemical properties, and putative function. J Biol Chem

1996;271:9550-9.

55. Kino T, Su YA, Chrousos GP. Human glucocorticoid receptor isoform beta: recent

understanding of its potential implications in physiology and pathophysiology.

Cell Mol Life Sci 2009;66:3435-48.

56. Lewis-Tuffin LJ, Cidlowski JA. The physiology of human glucocorticoid receptor

beta (hGRbeta) and glucocorticoid resistance. Ann N YAcad Sci 2006;1069:1-9.

57. Hauk PJ, Hamid QA, Chrousos GP, Leung DY. Induction of corticosteroid insen-

sitivity in human PBMCs by microbial superantigens. J Allergy Clin Immunol

2000;105:782-7.
























































































































































58. Tliba O, Cidlowski JA, Amrani Y. CD38 expression is insensitive to steroid action

in cells treated with tumor necrosis factor-alpha and interferon-gamma by a

mechanism involving the up-regulation of the glucocorticoid receptor beta iso-

form. Mol Pharmacol 2006;69:588-96.

59. Webster JC, Oakley RH, Jewell CM, Cidlowski JA. Proinflammatory cytokines

regulate human glucocorticoid receptor gene expression and lead to the accu-

mulation of the dominant negative beta isoform: a mechanism for the gener-

ation of glucocorticoid resistance. Proc Natl Acad Sci U S A 2001;98:

6865-70.

60. Varricchio L, Masselli E, Alfani E, Battistini A, Migliaccio G, Vannucchi AM,

et al. The dominant negative beta isoform of the glucocorticoid receptor is

uniquely expressed in erythroid cells expanded from polycythemia vera patients.

Blood 2011;118:425-36.

61. Jain A, Wordinger RJ, Yorio T, Clark AF. Spliceosome protein (SRp) regulation

of glucocorticoid receptor isoforms and glucocorticoid response in human trabec-

ular meshwork cells. Invest Ophthalmol Vis Sci 2012;53:857-66.

62. Xu Q, Leung DY, Kisich KO. Serine-arginine-rich protein p30 directs alternative

splicing of glucocorticoid receptor pre-mRNA to glucocorticoid receptor beta in

neutrophils. J Biol Chem 2003;278:27112-8.

63. Zhu J, Gong JY, Goodman OB Jr, Cartegni L, Nanus DM, Shen R. Bombesin at-

tenuates pre-mRNA splicing of glucocorticoid receptor by regulating the expres-

sion of serine-arginine protein p30c (SRp30c) in prostate cancer cells. Biochim

Biophys Acta 2007;1773:1087-94.

64. Derijk RH, Schaaf MJ, Turner G, Datson NA, Vreugdenhil E, Cidlowski J, et al.

A human glucocorticoid receptor gene variant that increases the stability of the

glucocorticoid receptor beta-isoform mRNA is associated with rheumatoid

arthritis. J Rheumatol 2001;28:2383-8.

65. Schaaf MJ, Cidlowski JA. AUUUA motifs in the 3’UTR of human glucocorticoid

receptor alpha and beta mRNA destabilize mRNA and decrease receptor protein

expression. Steroids 2002;67:627-36.

66. Geelhoed JJ, van Duijn C, van Osch-Gevers L, Steegers EA, Hofman A, Helbing

WA, et al. Glucocorticoid receptor-9beta polymorphism is associated with sys-

tolic blood pressure and heart growth during early childhood. The Generation

R Study. Early Hum Dev 2011;87:97-102.

67. Otte C, Wust S, Zhao S, Pawlikowska L, Kwok PY, Whooley MA. Glucocorticoid

receptor gene, low-grade inflammation, and heart failure: the Heart and Soul

study. J Clin Endocrinol Metab 2010;95:2885-91.

68. van den Akker EL, Koper JW, van Rossum EF, Dekker MJ, Russcher H, de Jong

FH, et al. Glucocorticoid receptor gene and risk of cardiovascular disease. Arch

Intern Med 2008;168:33-9.

69. Kino T, Manoli I, Kelkar S, Wang Y, Su YA, Chrousos GP. Glucocorticoid recep-

tor (GR) beta has intrinsic, GRalpha-independent transcriptional activity. Bio-

chem Biophys Res Commun 2009;381:671-5.

70. Lewis-Tuffin LJ, Jewell CM, Bienstock RJ, Collins JB, Cidlowski JA. The Human

glucocorticoid receptor b (hGRb) binds RU-486 and is transcriptionally active.

Mol Cell Biol 2007;27:2266-82.

71. Kelly A, Bowen H, Jee YK, Mahfiche N, Soh C, Lee T, et al. The glucocorticoid

receptor beta isoform can mediate transcriptional repression by recruiting histone

deacetylases. J Allergy Clin Immunol 2008;121:203-8.e1.

72. Kim SH, Kim DH, Lavender P, Seo JH, Kim YS, Park JS, et al. Repression of

TNF-alpha-induced IL-8 expression by the glucocorticoid receptor-beta involves

inhibition of histone H4 acetylation. Exp Mol Med 2009;41:297-306.

73. Li LB, Leung DY, Martin RJ, Goleva E. Inhibition of histone deacetylase 2

expression by elevated glucocorticoid receptor beta in steroid-resistant asthma.

Am J Respir Crit Care Med 2010;182:877-83.

74. DuBois DC, Sukumaran S, Jusko WJ, Almon RR. Evidence for a glucocorticoid

receptor beta splice variant in the rat and its physiological regulation in liver. Ste-

roids 2013;78:312-20.

75. Hinds TD Jr, Ramakrishnan S, Cash HA, Stechschulte LA, Heinrich G, Najjar

SM, et al. Discovery of glucocorticoid receptor-beta in mice with a role in meta-

bolism. Mol Endocrinol 2010;24:1715-27.

76. Schaaf MJ, Champagne D, van Laanen IH, van Wijk DC, Meijer AH, Meijer OC,

et al. Discovery of a functional glucocorticoid receptor beta-isoform in zebrafish.

Endocrinology 2008;149:1591-9.

77. Ray DW, Davis JR, White A, Clark AJ. Glucocorticoid receptor structure and

function in glucocorticoid-resistant small cell lung carcinoma cells. Cancer Res

1996;56:3276-80.

78. Beger C, Gerdes K, Lauten M, Tissing WJ, Fernandez-Munoz I, Schrappe M,

et al. Expression and structural analysis of glucocorticoid receptor isoform

gamma in human leukaemia cells using an isoform-specific real-time polymerase

chain reaction approach. Br J Haematol 2003;122:245-52.

79. Rivers C, Levy A, Hancock J, Lightman S, Norman M. Insertion of an amino acid

in the DNA-binding domain of the glucocorticoid receptor as a result of alterna-

tive splicing. J Clin Endocrinol Metab 1999;84:4283-6.

80. Moalli PA, Pillay S, Krett NL, Rosen ST. Alternatively spliced glucocorticoid re-

ceptor messenger RNAs in glucocorticoid-resistant human multiple myeloma

cells. Cancer Res 1993;53:3877-9.

81. de Lange P, Segeren CM, Koper JW, Wiemer E, Sonneveld P, Brinkmann AO,

et al. Expression in hematological malignancies of a glucocorticoid receptor

splice variant that augments glucocorticoid receptor-mediated effects in trans-

fected cells. Cancer Res 2001;61:3937-41.

82. Gaitan D, DeBold CR, Turney MK, Zhou P, Orth DN, Kovacs WJ. Glucocorticoid

receptor structure and function in an adrenocorticotropin-secreting small cell lung

cancer. Mol Endocrinol 1995;9:1193-201.

83. Krett NL, Pillay S, Moalli PA, Greipp PR, Rosen ST. A variant glucocorticoid re-

ceptor messenger RNA is expressed in multiple myeloma patients. Cancer Res

1995;55:2727-9.

84. Yudt MR, Jewell CM, Bienstock RJ, Cidlowski JA. Molecular origins for the

dominant negative function of human glucocorticoid receptor beta. Mol Cell

Biol 2003;23:4319-30.

85. Wu I, Shin SC, Cao Y, Bender IK, Jafari N, Feng G, et al. Selective glucocorticoid

receptor translational isoforms reveal glucocorticoid-induced apoptotic transcrip-

tomes. Cell Death Dis 2013;4:e453.

86. Gross KL, Oakley RH, Scoltock AB, Jewell CM, Cidlowski JA. Glucocorticoid

receptor a isoform-selective regulation of antiapoptotic genes in osteosarcoma

cells: a new mechanism for glucocorticoid resistance. Mol Endocrinol 2011;25:

1087-99.

87. Bender IK, Cao Y, Lu NZ. Determinants of the heightened activity of glucocor-

ticoid receptor translational isoforms. Mol Endocrinol 2013;27:1577-87.

88. Nehme A, Lobenhofer EK, Stamer WD, Edelman JL. Glucocorticoids with

different chemical structures but similar glucocorticoid receptor potency regulate

subsets of common and unique genes in human trabecular meshwork cells. BMC

Med Genomics 2009;2:58.

89. Cao Y, Bender IK, Konstantinidis AK, Shin SC, Jewell CM, Cidlowski JA, et al.

Glucocorticoid receptor translational isoforms underlie maturational stage-

specific glucocorticoid sensitivities of dendritic cells in mice and humans. Blood

2013;121:1553-62.

90. Wang SC, Myers S, Dooms C, Capon R, Muscat GE. An ERRbeta/gamma agonist

modulates GRalpha expression, and glucocorticoid responsive gene expression in

skeletal muscle cells. Mol Cell Endocrinol 2010;315:146-52.

91. Sinclair D, Webster MJ, Wong J, Weickert CS. Dynamic molecular and anatom-

ical changes in the glucocorticoid receptor in human cortical development. Mol

Psychiatry 2011;16:504-15.

92. Sinclair D, Webster MJ, Fullerton JM, Weickert CS. Glucocorticoid receptor

mRNA and protein isoform alterations in the orbitofrontal cortex in schizophrenia

and bipolar disorder. BMC Psychiatry 2012;12:84.

93. Sinclair D, Tsai SY, Woon HG, Weickert CS. Abnormal glucocorticoid receptor

mRNA and protein isoform expression in the prefrontal cortex in psychiatric

illness. Neuropsychopharmacology 2011;36:2698-709.

94. Pedersen KB, Geng CD, Vedeckis WV. Three mechanisms are involved in gluco-

corticoid receptor autoregulation in a human T-lymphoblast cell line. Biochem-

istry 2004;43:10851-8.

95. Russcher H, van Rossum EF, de Jong FH, Brinkmann AO, Lamberts SW, Koper

JW. Increased expression of the glucocorticoid receptor-A translational isoform

as a result of the ER22/23EK polymorphism. Mol Endocrinol 2005;19:1687-96.

96. Beck IM, Vanden Berghe W, Vermeulen L, Yamamoto KR, Haegeman G, De

Bosscher K. Crosstalk in inflammation: the interplay of glucocorticoid receptor-

based mechanisms and kinases and phosphatases. Endocr Rev 2009;30:830-82.

97. Galliher-Beckley AJ, Cidlowski JA. Emerging roles of glucocorticoid receptor

phosphorylation in modulating glucocorticoid hormone action in health and dis-

ease. IUBMB Life 2009;61:979-86.

98. Kumar R, Calhoun WJ. Differential regulation of the transcriptional activity of

the glucocorticoid receptor through site-specific phosphorylation. Biologics

2008;2:845-54.

99. Wang Z, Frederick J, Garabedian MJ. Deciphering the phosphorylation ‘‘code’’ of

the glucocorticoid receptor in vivo. J Biol Chem 2002;277:26573-80.

100. Avenant C, Ronacher K, Stubsrud E, Louw A, Hapgood JP. Role of ligand-

dependent GR phosphorylation and half-life in determination of ligand-specific

transcriptional activity. Mol Cell Endocrinol 2010;327:72-88.

101. Galliher-Beckley AJ, Williams JG, Cidlowski JA. Ligand-independent phosphor-

ylation of the glucocorticoid receptor integrates cellular stress pathways with nu-

clear receptor signaling. Mol Cell Biol 2011;31:4663-75.

102. Webster JC, Jewell CM, Bodwell JE, Munck A, Sar M, Cidlowski JA. Mouse

glucocorticoid receptor phosphorylation status influences multiple functions of

the receptor protein. J Biol Chem 1997;272:9287-93.

103. Blind RD, Garabedian MJ. Differential recruitment of glucocorticoid receptor

phospho-isoforms to glucocorticoid-induced genes. J Steroid Biochem Mol Biol

2008;109:150-7.
























































































































































NOVEMBER 2013


104. Chen W, Dang T, Blind RD, Wang Z, Cavasotto CN, Hittelman AB, et al. Gluco-

corticoid receptor phosphorylation differentially affects target gene expression.

Mol Endocrinol 2008;22:1754-66.

105. Miller AL, Garza AS, Johnson BH, Thompson EB. Pathway interactions between

MAPKs, mTOR, PKA, and the glucocorticoid receptor in lymphoid cells. Cancer

Cell Int 2007;7:3.

106. Miller AL, Webb MS, Copik AJ, Wang Y, Johnson BH, Kumar R, et al. p38

Mitogen-activated protein kinase (MAPK) is a key mediator in glucocorticoid-

induced apoptosis of lymphoid cells: correlation between p38 MAPK activation

and site-specific phosphorylation of the human glucocorticoid receptor at serine

211. Mol Endocrinol 2005;19:1569-83.

107. Simic I, Maric NP, Mitic M, Soldatovic I, Pavlovic Z, Mihaljevic M, et al. Phos-

phorylation of leukocyte glucocorticoid receptor in patients with current episode

of major depressive disorder. Prog Neuropsychopharmacol Biol Psychiatry 2013;

40:281-5.

108. Galliher-Beckley AJ, Williams JG, Collins JB, Cidlowski JA. Glycogen synthase

kinase 3beta-mediated serine phosphorylation of the human glucocorticoid recep-

tor redirects gene expression profiles. Mol Cell Biol 2008;28:7309-22.

109. Ismaili N, Blind R, Garabedian MJ. Stabilization of the unliganded glucocorticoid

receptor by TSG101. J Biol Chem 2005;280:11120-6.

110. Itoh M, Adachi M, Yasui H, Takekawa M, Tanaka H, Imai K. Nuclear export of

glucocorticoid receptor is enhanced by c-Jun N-terminal kinase-mediated phos-

phorylation. Mol Endocrinol 2002;16:2382-92.

111. Deroo BJ, Rentsch C, Sampath S, Young J, DeFranco DB, Archer TK. Proteaso-

mal inhibition enhances glucocorticoid receptor transactivation and alters its sub-

nuclear trafficking. Mol Cell Biol 2002;22:4113-23.

112. Wallace AD, Cidlowski JA. Proteasome-mediated glucocorticoid receptor degra-

dation restricts transcriptional signaling by glucocorticoids. J Biol Chem 2001;

276:42714-21.

113. Wang X, DeFranco DB. Alternative effects of the ubiquitin-proteasome pathway

on glucocorticoid receptor down-regulation and transactivation are mediated by

CHIP, an E3 ligase. Mol Endocrinol 2005;19:1474-82.

114. Druker J, Liberman AC, Antunica-Noguerol M, Gerez J, Paez-Pereda M, Rein T,

et al. RSUME enhances glucocorticoid receptor SUMOylation and transcriptional

activity. Mol Cell Biol 2013;33:2116-27.

115. Davies L, Karthikeyan N, Lynch JT, Sial EA, Gkourtsa A, Demonacos C, et al.

Cross talk of signaling pathways in the regulation of the glucocorticoid receptor

function. Mol Endocrinol 2008;22:1331-44.

116. Holmstrom S, Van Antwerp ME, Iniguez-Lluhi JA. Direct and distinguishable

inhibitory roles for SUMO isoforms in the control of transcriptional synergy.

Proc Natl Acad Sci U S A 2003;100:15758-63.

117. Holmstrom SR, Chupreta S, So AY, Iniguez-Lluhi JA. Sumo-mediated inhibition

of glucocorticoid receptor synergistic activity depends on stable assembly at the

promoter but not on DAXX. Mol Endocrinol 2008;22:2061-75.

118. Le Drean Y, Mincheneau N, Le Goff P, Michel D. Potentiation of glucocorticoid

receptor transcriptional activity by sumoylation. Endocrinology 2002;143:

3482-9.

119. Tian S, Poukka H, Palvimo JJ, Janne OA. Small ubiquitin-related modifier-1

(SUMO-1) modification of the glucocorticoid receptor. Biochem J 2002;367:

907-11.

120. Ito K, Yamamura S, Essilfie-Quaye S, Cosio B, Ito M, Barnes PJ, et al. Histone

deacetylase 2-mediated deacetylation of the glucocorticoid receptor enables

NF-kappaB suppression. J Exp Med 2006;203:7-13.

121. Charmandari E, Chrousos GP, Lambrou GI, Pavlaki A, Koide H, Ng SS, et al.

Peripheral CLOCK regulates target-tissue glucocorticoid receptor transcriptional

activity in a circadian fashion in man. PLoS One 2011;6:e25612.

122. Fan BJ, Wang DY, Tham CC, Lam DS, Pang CP. Gene expression profiles of hu-

man trabecular meshwork cells induced by triamcinolone and dexamethasone.

Invest Ophthalmol Vis Sci 2008;49:1886-97.

123. Adcock IM, Nasuhara Y, Stevens DA, Barnes PJ. Ligand-induced differentiation

of glucocorticoid receptor (GR) trans-repression and transactivation: preferential

targetting of NF-kappaB and lack of I-kappaB involvement. Br J Pharmacol 1999;

127:1003-11.

124. Chen SH, Masuno K, Cooper SB, Yamamoto KR. Incoherent feed-forward regu-

latory logic underpinning glucocorticoid receptor action. Proc Natl Acad Sci U S

A 2013;110:1964-9.

125. Barnes PJ. Corticosteroid resistance in patients with asthma and chronic obstruc-

tive pulmonary disease. J Allergy Clin Immunol 2013;131:636-45.

126. Yang N, Ray DW, Matthews LC. Current concepts in glucocorticoid resistance.

Steroids 2012;77:1041-9.

127. Clark AR, Belvisi MG. Maps and legends: the quest for dissociated ligands of the

glucocorticoid receptor. Pharmacol Ther 2012;134:54-67.

128. Vandevyver S, Dejager L, Tuckermann J, Libert C. New insights into the anti-

inflammatory mechanisms of glucocorticoids: an emerging role for

glucocorticoid-receptor-mediated transactivation. Endocrinology 2013;154:

993-1007.













































































Reproduced with permission of the copyright owner. Further reproduction prohibited withoutpermission.

Ligand-induced di�erentiation of glucocorticoid receptor (GR)trans-repression and transactivation: preferential targetting ofNF-kB and lack of I-kB involvement

*,1Ian M. Adcock, 1Yasuyuki Nasuhara, 1David A. Stevens & 1Peter J. Barnes

1Thoracic Medicine, Imperial College School of Medicine at NHLI, Dovehouse Street, London SW3 6LY

1 Glucocorticoids are highly e�ective in controlling chronic in¯ammatory diseases, such as asthmaand rheumatoid arthritis, but the exact molecular mechanism of their anti-in¯ammatory actionremains uncertain. They act by binding to a cytosolic receptor (GR) resulting in activation orrepression of gene expression. This may occur via direct binding of the GR to DNA(transactivation) or by inhibition of the activity of transcription factors such as AP-1 and NF-kB(transrepression).

2 The topically active steroids ¯uticasone propionate (EC50=1.8610711M) and budesonide

(EC50=5.0610711M) were more potent in inhibiting GM-CSF release from A549 cells than

tipredane (EC50=8.3610710M), butixicort (EC50=3.761078

M) and dexamethasone(EC50=2.261079

M). The anti-glucocorticoid RU486 also inhibited GM-CSF release in these cells(IC50=1.8610710

M).

3 The concentration-dependent ability of ¯uticasone propionate (EC50=9.8610710M), budesonide

(EC50=1.161079M) and dexamethasone (EC50=3.661078

M) to induce transcription of the b2-receptor was found to correlate with GR DNA binding and occurred at 10 ± 100 fold higherconcentrations than the inhibition of GM-CSF release. No induction of the endogenous inhibitorsof NF-kB, IkBa or I-kBb, was seen at 24 h and the ability of IL-1b to degrade and subsequentlyinduce IkBa was not altered by glucocorticoids.

4 The ability of ¯uticasone propionate (IC50=0.5610711M), budesonide (IC50=2.7610711

M),dexamethasone (IC50=0.561079

M) and RU486 (IC50=2.7610711M) to inhibit a 36kB was

associated with inhibition of GM-CSF release.

5 These data suggest that the anti-in¯ammatory properties of a range of glucocorticoids relate totheir ability to transrepress rather than transactivate genes.

Keywords: Glucocorticoids; in¯ammation; GM-CSF; NF-kB; cross-coupling; RU486

Abbreviations: AP-I, activator protein-1; Bud, budesonide; Dex, dexamethasone; ECL, enhanced chemiluminescence; FCS,foetal calf serum; FP, ¯uticasone propionate; GCs, glucocorticoids; GM-CSF, granulocyte-macrophage colonystimulating factor; GR, glucocorticoid receptor; I-kBa, inhibitor of NF-kB; NF-kB, nuclear factor-kB; P/CAF,p300/CBP associated protein; PAO, phenylarsine oxide; PBS, phosphate bu�ered saline; PCIP, p300/CBP co-integrator protein; SRC-1, steroid-receptor coactivator-1; STAT, signal transduction and activation oftranscription; TNFa, tumour necrosis factor-a; TRE, TPA-response element

Introduction

Cytokines, such as IL-1b, tumour necrosis factor-a (TNFa)and granulocyte-macrophage colony stimulating factor (GM-CSF), are released in a co-ordinate network and play animportant role in chronic in¯ammation. As such, the pattern

of cytokine expression largely determines the nature andpersistence of the in¯ammatory response (Barnes & Adcock,1993). Cytokines produce their cellular e�ects by activation ofvarious transcription factors such as activator protein-1 (AP-

I), nuclear factor-kB (NF-kB), and the signal transduction andactivation of transcription (STAT) family. Furthermore, theexpression of many of these cytokines and their receptors are

also upregulated by these transcription factors. The increasedexpression of some of these factors may be responsible for theprolonged in¯ammation seen in asthma. AP-1 and NF-kB can

also be induce, and be induced by, numerous other mediatorssuch as NO, histamine, and eicosanoids (Barnes & Adcock,1993).

Glucocorticoids (GCs) are the most e�ective anti-in¯am-

matory therapy for the treatment of asthma and act byreducing airway hyper-responsiveness and suppressing theairway in¯ammatory response (Barnes, 1995). However, their

exact mechanisms and cellular targets in the lung areuncertain. Glucocorticoid receptors (GR) are predominantlylocalized to the airway epithelium and endothelium (Adcocket al., 1996), therefore, these are likely to be important sites

of the anti-in¯ammatory action of steroids, especially whendelivered by the inhaled route. Airway epithelial cells do notact solely as a physical barrier but act as important regulators

of the in¯ammatory reaction, responding to variousin¯ammatory mediators by the production of a wide rangeof cytokines, chemokines and other in¯ammatory mediators

(Levine, 1995).Classically GCs act by binding to, and activating, the

cytosolic GR. Upon activation the GR dimerizes and

translocates to the nucleus. Within the nucleus GR binds tospeci®c DNA elements (GREs) in the promoters of responsivegenes resulting in modulation of transcription (Beato et al.,1996; Strahle et al., 1988). Important genes induced by*Author for correspondence; E-mail: [email protected]

British Journal of Pharmacology (1999) 127, 1003 ± 1011 ã 1999 Stockton Press All rights reserved 0007 ± 1188/99 $12.00

http://www.stockton-press.co.uk/bjp

glucocorticoids include the b2-receptor (Collins et al., 1988)and lipocortin-1 (Flower & Rothwell, 1994). Lipocortininhibits the activation of cytosolic phospholipase A2 and thus

may inhibit the synthesis of leukotrienes and prostanoids,although there is some doubt whether this mechanism isimportant in inhibiting asthmatic in¯ammation (Davidson etal., 1990).

The mechanisms involved in GR-mediated gene repressionare less well understood. Experiments involving overexpressionof the subunits of AP-1 and NF-kB along with deletion

mutants of speci®c DNA binding moieties have indicated thatthe predominant mechanism of glucocorticoid down-regula-tion of in¯ammatory genes requires a direct protein-protein

interaction with these, or related, transcription factors (Karin,1998). An additional mechanism of repression by glucocorti-coids on NF-kB mediated transcription has been reported in

cultured monocytes and T-lymphocytes (Auphan et al., 1995;Scheinman et al., 1995a). Following steroid treatment a rapidinduction of the inhibitor of NF-kB (I-kBa) mRNA andprotein synthesis occurs. Newly synthesized I-kBa interacts

with NF-kB heterodimers within the cytoplasm, and possiblythe nucleus (Zabel et al., 1993) thereby inhibiting NF-kB DNAbinding and transcriptional activation by cytokines. This

mechanism does not appear to occur in all cells examined(Brostjan et al., 1996; Heck et al., 1997).

We have, therefore, investigated the role of these various

mechanisms in mediating the ability of the glucocorticoids toinhibit IL-1b-induced release of GM-CSF, an NFkB-induciblegene (Kochetkova & Shannon, 1996), from a human lung

epithelial-like cell line (A549). We have also examined the GR-induced inhibition of NFkB and AP-1 activity and the abilityof GR to bind to its DNA binding motif and induce the b2-receptor and IkBa and b proteins. In addition, we have

examined the relative abilities of a number of topically actingsteroids to activate or repress gene transcription to investigatewhether current steroids can di�erentiate between these two

modes of glucocorticoid action.

Methods

Drugs and chemicals

Fluticasone propionate, budesonide, tipredane and butixicortwere kindly supplied by Dr M. Johnson (Glaxo-Wellcome,U.K.). RU486 (Mifepristone) was supplied by Dr A. Phillibert

(Rousel-Uclaf, France). Enzymes were obtained from Promega(Cambridge, U.K.) and all other reagents, except where stated,were obtained from Sigma (Dorset, U.K.).

Cell culture

A549 cells were grown to con¯uence in Dulbecco's modi®edmedium containing 10% foetal calf serum (FCS) beforeincubation for 72 h in serum-free media, as previouslydescribed (Newton et al., 1996). Cells were then used for the

analysis of glucocorticoid action on GM-CSF release, b2-receptor, I-kB expression and transcription factor activity.

GM-CSF release

A549 cells were cultured for 24 h with 1 ng ml71 IL-1b in the

presence or absence of various concentrations of steroids.After this time the culture supernatant was removed and storedat 7708C until samples were analysed. GM-CSF concentra-tions were measured using a speci®c ELISA calibrated with

human recombinant GM-CSF (0 ± 200 pg ml71, PharMingen,Lugano, Switzerland). Coating solution (0.1 M NaHCO3,pH 8.2) containing 2 mg ml71 rat anti human GM-CSF

monoclonal antibody was used to coat an enhanced proteinbinding ELISA plate and incubated overnight at 48C. Theplate was washed three times in 0.05% Tween-20 in phosphatebu�ered saline (PBS) before blocking with 10% FCS in PBS

for 2 h at 188C. Plates were washed again and incubated withsamples and standards (diluted in 10% serum in PBS)overnight at 48C before washing again with 0.05% Tween-20

in PBS. 1 mg ml71 biotinylated rat anti human GM-CSFdiluted in 0.05% Tween-20 in PBS was incubated for 45 min at188C before extensive washing. The signal was detected

following a 30 min incubation at 188C with a 1 : 400 dilutionof 1 mg ml71 avidin-peroxidase solution (Sigma), extensivewashing and ®nal addition of ABTS substrate solution

containing 1 ml ml71 H2O2. The colour reaction products wereread at 405 nm.

Electrophoretic mobility shift assay

Nuclear and cytosolic proteins were extracted from A549 cellsas previously described (Adcock et al., 1995). Brie¯y, cells were

collected and lysed in 200 ml of Bu�er A 10 (mM) HEPESMgCl2 1.5, KCl 10, DTT 0.5, 0.1% Nonidet P40, andincubated at 48C for 15 min. After microcentrifugation for

10 s and collection of the cytoplasmic fraction, the nuclearpellet was lysed with 20 ml of Bu�er B (mM) HEPES 20, MgCl21.5, NaCl 0.42, DTT 0.5, 25% glycerol, PMS 0.5, EDTA 0.2.

The subsequent soluble fraction was mixed with 100 ml ofbu�er C (mM) HEPES 20, KCl 50, DTT 0.5, PMSF 0.5,EDTA 0.2. 2 mg nuclear protein from each sample was pre-incubated at 48C for 30 min in binding bu�er (mM) Tris HCl

10, pH 7.5 MgCl2 1, EDTA 0.5, DTT 0.5, NaCl 50, 4%glycerol, 0.1 mg ml71 salmon sperm DNA). Double-strandedoligonucleotides encoding the speci®c sequence of GRE (5'-TCGACTGTACAGGATGTTCTAGCTACT-) (Promega)were end-labelled with [g732P]-ATP and T4 polynucleotidekinase. Each sample was then incubated with 50,000 c.p.m. of

labelled oligonucleotide for 40 min at 48C. Protein-DNAcomplexes were separated on a 6% polyacrylamide gel using0.256Tris-Borate-EDTA running bu�er. Speci®city wasdetermined by the addition of excess unlabelled double

stranded oligonucleotides.

Western blotting

Total cellular proteins were extracted from A549 cells byfreeze-thawing samples in lysis bu�er (mM) HEPES 20,

MgCl2 1.5, NaCl 0.42, DTT 0.5, 25% glycerol, PMSF 0.5,EDTA 0.2 (Adcock et al., 1996). 30 ± 50 mg total solubleprotein extracts were size fractionated on 10% PAGE gels

and transblotted onto Nitrocellulose-ECL membranes (Amer-sham International, Amersham, U.K.). Membranes wereblocked overnight with 2% casein prior to incubation witheither 1 : 1000 rabbit anti-human I-kBa antibody (Santa Cruz

Biotechnology Inc., Devizes, U.K.), 1 : 1000 rabbit anti-human I-kBb antibody (Santa Cruz Biotechnology Inc) or1 : 600 rabbit anti-human human b2 receptor (kindly donated

by Dr J. MacDermot, RPMS, U.K.) at 188C for 3 h. Afterwashing (3620 min in PBS-Tween), bound antibody wasdetected using 1 : 7000 sheep anti-rabbit antibody (F(ab')2fragment) linked to horseradish peroxidase (AmershamInternational, U.K.) and bound complexes detected usingEnhanced Chemiluminescence (ECL, Amersham Interna-tional) (Adcock et al., 1996).

No I-kB involvement in glucocorticoid repression1004 I.M. Adcock et al

Immunoprecipitation

Cells were treated for 30 min with 1 mM dexamethasone or

¯uticasone propionate in the presence of 1 ng ml71 IL-1b.After incubation the cells were washed three times with freshmedia before extraction in Bu�er A above and two cycles offreeze-thaw. At the end of this time, soluble proteins (100 mg)were incubated for 18 h at 48C with either an anti-human p65antibody, an anti-human GR antibody or pre-immune serum(Santa Cruz). Immune complexes were precipitated with

protein A-sepharose and loaded onto 10% PAGE gels.Following electrophoretic separation, proteins were electro-blotted and transferred to ECL-nitrocellulose membranes and

probed for the presence of p65 or GR.

Transient transfection and luciferase assay

Sense and antisense oligonucleotides encoding 6 copies of theconsensus DNA binding site for AP-1 (TRE, 5'-CGCTTGAT-GAGTCAGCCGGAA-) were annealed by allowing to cool

slowly to room temperature after incubation at 958C for 2 min.The double stranded oligonucleotide was then inserted intopGL3-Basic at the SmaI site. The number of incorporated

TRE sites and the direction of incorporation were con®rmedby sequencing.

Cells were grown to con¯uence and then treated for 2 days

in serum-free media. pGM-CSF(7123)-Luc (containing the7123 to +1 sequence of the human GM-CSF promoter),pGL-3kBLuc or pGL-66TRE-Luc were incubated with 2.5 mlTfx750 reagent (Promega) mg71 DNA ml71 of serum freemedia for 15 min at room temperature. Cells were transfectedby the addition of 1 ml of media containing DNA-Tfx750(5 ml Tfx750 mg71 DNA) for 1.5 h before washing in fresh

media and incubation in 1 ml serum free media. All cells weretransfected with 1 mg pSV-b-gal vector (Promega) to controlfor transfection e�ciency. After 18 h the media was changed

and cells stimulated with varying concentrations of glucocorti-coids in the presence or absence of 1 ng ml71 of IL-1b. Cellswere harvested by scraping and resuspended in 16reporter

Lysis Bu�er (Promega). After incubation at room temperaturefor 15 min lysates were vortexed for 10 s and subjected to onefreeze-thaw cycle. Cellular debris was pelleted and total proteinwas measured. Luciferase assays were performed using 20 ml ofextract and 50 ml Luciferase Assay Reagent (Promega) andluminescence measured with a TD 20/20 Luminometer (TurnerDesigns, Hemel Hempstead, U.K.). Relative luminescence

readings were normalized to b-galactosidase expression andexpressed as a percentage of activation relative to control orIL-1b-stimulated release.

Results

E�ects of glucocorticoids on IL-1b-stimulated GM-CSFrelease

IL-1b caused a concentration-dependent increase in GM-CSFrelease after 24 h in A549 cells (EC50=0.3 ng ml71). IL-1b(1 ng ml71) stimulated the production of approximately 80%

maximal levels of GM-CSF (22.5+4.5 ng ml71). No detect-able GM-CSF was found in the supernatant of controluntreated cells. Co-incubation with ¯uticasone propionate,

budesonide or dexamethasone all produced a concentration-dependent inhibition of IL-1b-stimulated GM-CSF releaseover this time period. Fluticasone propionate(EC50=1.8610711

M) and budesonide (EC50=5.0610711M)

produced a 122 fold and a 44 fold greater inhibition of GM-CSF release respectively than dexamethasone(EC50=2.261079

M) (Figure 1a and Table 1). The anti-

glucocorticoid RU486 also caused a concentration-dependentinhibition of IL-1b-stimulated GM-CSF release from A549cells with an EC50=1.8610710

M. At concentrations greaterthan 1077

M RU486 was less e�ective at inhibiting IL-1b-induced GM-CSF release (Figure 1b). Furthermore, the abilityof 10710

M ¯uticasone propionate, budesonide or dexametha-sone to inhibit IL-1b-stimulated GM-CSF release was not

inhibited by 1078M RU486 (Figure 1c). Both tipredane

(EC50=8.3610710M) and butixicort (EC50=3.761078

M)also caused a concentration-dependent inhibition of IL-1b-stimulated GM-CSF release. The ability of IL-1b to stimulateGM-CSF release was markedly attenuated by the NF-kBinhibitor phenylarsine oxide (PAO) in a concentration-

dependent manner with a maximal e�ect at 10 mM (Figure1d). The proteosome inhibitor CBZ-leucine-leucine-leucinal(LLLal) caused a concentration-dependent inhibition of IL-1b(1 ng ml71)-stimulated GM-CSF release (EC50=30 mM).

E�ect of glucocorticoids on DNA binding activity

The ability of ¯uticasone propionate and dexamethasone tostimulate DNA binding was assessed by EMSA. Fluticasonepropionate and dexamethasone produced a concentration-

dependent increase in DNA binding. Fluticasone propionate(EC50=5.261079

M) and budesonide (EC50=6.461079M)

caused a marked increase in DNA binding with an

approximately 10 fold higher potency than that seen fordexamethasone (EC50=4.661078

M) (Table 1). Moreoverthe maximal level of DNA binding induced by ¯uticasonepropionate was greater (30 fold stimulation) than that seen

with dexamethasone (5 fold stimulation). At higherconcentrations of ¯uticasone propionate (1076

M) theinducibility of DNA binding was much reduced possibly

re¯ecting the relative partial agonist activity of this drug.RU486 gave no consistent increase in DNA binding (Figure2). Con®rmation that the correct band was detected was

obtained by using an excess of unlabelled oligonucleotideand also by the use of a speci®c GR antibody thatsupershifted the retarded band.

E�ects of glucocorticoids on the induction of theb2-adrenoceptor

In order to ensure that the glucocorticoids were able to inducegene expression in these cells the ability of these glucocorti-coids to induce the expression of the b2-adrenoceptor after

24 h was investigated by Western blotting (Figure 3). Theconcentration-dependent ability of ¯uticasone propionate(EC50=1.061079

M), budesonide (EC50=1.161079M) and

dexamethasone (EC50=3.661078M) to induce a 2 ± 3 fold

increase in the expression of the b2-receptor at 24 h correlatedwith the induction of GR DNA binding. This occurred at 10 ±100 fold higher concentrations than that which repressed IL-

1b-stimulated GM-CSF release (Figure 3). Fluticasonepropionate caused a greater increase in b2-receptor expressionthan either budesonide or dexamethasone. RU486 had no

e�ect on b2-adrenoceptor expression at any time or at anyconcentration tested (Table 1).

E�ects of glucocorticoids on I-kBa and b expression

The ability of glucocorticoids to a�ect the induction of thecytoplasmic inhibitor of NF-kB, I-kBa, was also investi-

No I-kB involvement in glucocorticoid repression 1005I.M. Adcock et al

Figure 1 (a) Concentration-dependent inhibition of interleukin (IL)-1b (1 ng ml71)-stimulated granulocyte-macrophage colonystimulating factor (GM-CSF) release into the media from A549 cells at 24 h following ¯uticasone propionate (FP), budesonide (Bud)and dexamethasone (Dex) treatment. (b) Concentration-dependent inhibition of IL-1b (1 ng ml71)-stimulated GM-CSF release fromA549 cells at 24 h following treatment with the anti-glucocorticoid RU486. (c) The e�ect of low concentration (1079

M) RU486 (RU)treatment on the inhibition of IL-1b-stimulated GM-CSF release by 10710

M FP, Bud and Dex. (d) The e�ects of increasingconcentrations of phenylarsine oxide (PAO) on IL-1b (1 ng ml71)-stimulated induction of GM-CSF release into culture medium at24 h. Results are plotted as the means+s.e.means of the percentage of maximal IL-1b-stimulated GM-CSF release in the absence anydrug. n=4±7 for each data point except in (c) where results are the mean of two independent experiments.

Table 1 Glucocorticoid e�ects on transactivation and transrepression in A549 cells

FP IC50 Bud IC50 Dex IC50 RU486 IC50

GM-CSF release*b2 ReceptorGRE BindingI-kBaI-kBa degradationI-kBbI-kBb degradationkB activity (*stim)kB activity (basal)TRE activity (*stim)TRE activity (basal)GM-CSF promotor*

1.8610711M

1.061079M

5.061079M

No inductionNo e�ect

No inductionNo e�ect1.8610711

M

0.5610711M

1.7610710M

1.1610710M

0.6610711M

5.0610711M

1.161079M

2.461079M



M

2.7610711M

ND1.0610710

M

ND

2.261079M

3.261078M

4.661078M



M

0.561079M

0.961079M

0.361079M

1.361079M

1.8610710M

No inductionNo inductionNo inductionNo e�ect


M

2.5610711M

ND7.1610711

M

ND

FP, ¯uticasone propionate; Bud, budesonide; Dex, dexamethasone; GM-CSF, granulocyte macrophage colony stimulating factor; GREbinding, glucocorticoid receptor DNA binding. *After stimulation with interleukin (IL)-1b (1 ng ml71). ND, experiment not performed.


gated in these cells by Western blot analysis. Cells wereincubated with glucocorticoids (10712 ± 1076

M) for varioustime periods of up to 24 h. None of the glucocorticoids

investigated (¯uticasone propionate, budesonide, dexametha-sone or RU486) had any e�ect on the expression of I-kBaprotein at any concentration tested in these cells (Figure4a). IL-1b (1 ng ml71) stimulation caused a rapid phos-

phorylation and subsequent degradation of I-kBa whichwas inhibited by the proteosome inhibitor CBZ-leucine-leucine-leucinal (LLLal, 50 mM) (Figure 4b). This was

followed at 60 ± 90 min by the induction of de novosynthesized I-kBa (Figure 4c). We further examined thee�ects of these glucocorticoids on this degradation and

reappearance of I-kBa within the cytoplasm of these cells.IL-1b caused a total loss of I-kBa protein from thecytoplasm within 2 ± 5 min. The expression of I-kBa protein

returned to control levels between 90 and 120 min. None ofthe drugs tested had any e�ect on the time course of I-kBadegradation or synthesis (Figure 4c). Glucocorticoids maya�ect NF-kB activation by altering the level of the I-kBassociated with longer term induction of NF-kB, I-kBb inthese cells. None of the glucocorticoids tested had anye�ect on the induction of I-kBb at 24 h. In contrast to I-

kBa, IL-1b did not cause a rapid degradation of I-kBb but

a

b

Figure 2 (a) Representative electrophoretic mobility shift assayshowing the concentration-dependent e�ect of ¯uticasone propionate(FP), Budesonide (Bud) and dexamethasone (Dex) on glucocorticoidreceptor (GR)-induced activation as represented by increased DNAbinding (GRE binding) (arrowed) within the nucleus after 2 hincubation. (b) Supershift assay of dexamethasone (1076

M)-stimulated GR DNA binding. Increased DNA binding is seenfollowing dexamethasone treatment (lane 2). Pre-incubation ofretarded complexes with an anti-GR antibody (lane 3) shows speci®cenhanced retardation of GR/GRE band. Speci®city of binding wasindicated by the addition of 100 fold excess unlabelled oligonucleo-tide (lane 4). Unbound oligonucleotide is indicated by an arrow atthe bottom of the gel. (c) Densitometric analysis of the retardedbands in (a) and corrected for maximal band intensity showing theconcentration-dependent increase in GRE binding following 2 hincubation with FP, Bud and Dex within the nucleus as a percentageof the maximal increase observed.

a

Figure 3 (a) Western blot analysis of b2-receptor (b2R) expression at24 h following increasing concentrations of ¯uticasone propionate(FP), budesonide (Bud), dexamethasone (dex) or RU486. The single47 kD band representing the b2-receptor is indicated by the arrow.Incubation with control media does not a�ect b2-receptor expression.(b) Graphical representation of the results shown in (a). Results areshown as the percentage change in b2-receptor band densitycompared to control untreated cells and are representative of fourindividual experiments and are reported as the means+s.e.means.


caused a decrease in I-kBb levels at 4 ± 6 h before returningto control levels which was not a�ected by glucocorticoids(Figure 4d and Table 1).

GR/p65 immune complex

Immunoprecipitation studies showed that the p65 subunit of

NF-kB was associated with GR, either directly or within acomplex, within A549 cells. Western blot analysis of theimmunoprecipitated p65 complex showed the presence of aspeci®c GR band. In the reverse experiment the immunopre-

cipitated GR complex showed the presence of p65 (Figure 5).In contrast, immunoprecipitation with pre-immune serumshowed no binding of GR or p65.

NF-kB- and TRE-driven reporter gene constructs

IL-1b (1 ng ml71) stimulation produced a signi®cant 3 foldinduction in luciferase activity with both the pGM-CSF(7123)-Luc and the pGL-3kBLuc vectors. IL-1b-stimulated 36kB-activated luciferase activity was inhibited in a concentration-dependent manner by ¯uticasone propionate (FP,IC50=1.8610711

M), budesonide (Bud, IC50=2.76 10711M),

dexamethasone (Dex, IC50=0.861079M) and RU486 (IC50=

8610711M) (Figure 6a). The basal expression of pGL-3kB-Luc

was also modulated by glucocorticoids. Fluticasone propionate(IC50=0.5610711

M), budesonide (IC50=2.7610711M) and

dexamethasone (IC50=0.561079M) caused a 50 ± 70% de-

crease in luciferase activity compared to those seen in controlunstimulated cells. RU486 also caused a concentration-

dependent inhibition of kB-driven luciferase activity (IC50=2.5610711

M) but themaximal reduction seenwas 50%of basallevels. IL-1b-stimulated GM-CSF-promoter driven luciferase

activity was inhibited in a concentration-dependent manner by¯uticasone propionate (IC50=0.6610711

M) and dexametha-sone (IC50=1.361079

M) (see Table 1).IL-1b (1 ng ml71) caused a 2.4 fold increase in a 66TRE

promoter-driven luciferase activity after 18 h of stimulation.IL-1b-stimulated TRE-promoter driven luciferase activity wasinhibited in a concentration-dependent manner by ¯uticasone

propionate (IC50=1.7610711M) and dexamethasone

(IC50=0.961079M) (Figure 6b). These glucocorticoids also

reduced basal expression of TRE activity, ¯uticasone

propionate (IC50=1.1610710M), budesonide (IC50=1.06

10710M) and dexamethasone (IC50=0.361079

M) all reducinglevels to less than 50% of those seen in control unstimulatedcells. RU486 also caused a concentration-dependent inhibition

of TRE activity (IC50=7.1610711M) but the maximal

reduction seen was 50% of basal levels. Inhibition of TREactivity occurred at approximately 10 fold lower concentra-

tions than that which caused inhibition of kB-driven luciferaseactivity and GM-CSF release (see Table 1).

Discussion

Fluticasone propionate and budesonide were more potent asinhibitors of GM-CSF release and NF-kB activity thandexamethasone. Although all these ligands were acting throughthe same receptor, ¯uticasone propionate and budesonide were

approximately ®ve times more potent at these targets thanwould be predicted purely by ligand binding a�nity. Incontrast, the ability of these drugs to modulate AP-1 reporter

gene activity was more closely related to their GR ligandbinding a�nity (see Brattsand & Linden (1996)). This suggeststhat altered conformational changes in the GR monomer may

alter the ability to repress gene expression in a ligand dependentmanner. This was further suggested by the ability of lowconcentration, but not high concentration, RU486 to suppressIL-1b-stimulated GM-CSF release and kB activity.

a

b

c

d

Figure 4 Western blot analysis of the time course of I-kBaexpression following 24 h treatment with various concentrations of¯uticasone propionate (FP), budesonide (Bud) and dexamethasone(Dex). Concentrations are reported as 7log of Molar concentrations.(b) Western blot analysis of the time course of I-kBa expressionfollowing up to 90 min treatment with IL-1b (1 ng ml71) in thepresence and absence of the proteosome inhibitor CBZ-leucine-leucine-leucinal (LLLal) (50 mM) I-kBa is indicated by the arrow. (c)Western blot analysis of the time course of I-kBa expressionfollowing up to 90 min treatment with IL-1b (1 ng ml71). The e�ectsof ¯uticasone propionate (FP, 10710

M), budesonide (Bud, 10710M)

and dexamethasone (Dex, 1079M) on IL-1b-stimulated I-kBa

phosphorylation, degradation and subsequent induction are shown.I-kBa and the slower migrating phosphorylated form of I-kBa areindicated by the arrows. (d) The lack of e�ect of ¯uticasonepropionate (FP, 10710

M) budesonide (Bud, 10710M) and dexa-

methasone (Dex, 1079M) on IL-1b-stimulated I-kBb phosphoryla-

tion, degradation and subsequent induction over 0 ± 8 h are shown.Results are representative of four individual experiments.


Figure 5 Western blot analysis of immunoprecipitated p65 and GR complexes. Cells were treated for 30 min with a combination ofIL-1b (1 ng ml71) and dexamethasone (1076

M) before cell were lysis. Total cell extracts were immunoprecipitated with an anti-human p65 antibody (lane 1), an anti-human GR antibody (lane 4) or with pre-immune serum (lanes 2 and 3) before separation by10% PAGE and detection of bands by either anti-human GR antibody (lanes 1 and 2) or anti-human p65 antibody (lanes 3 and 4).The speci®c GR or p65 bands are indicated arrows. The 50 kDa IgG heavy chain is detected in all samples and is also arrowed.Molecular weight markers are as indicated. The results are representative of three independent experiments.

Figure 6 (a) Inhibition of an IL-1b (1 ng ml71)-stimulated 36kB-driven luciferase reporter gene by ¯uticasone propionate (FP),budesonide (Bud), dexamethasone (Dex) and RU486. Results are expressed as relative light units/unit b-galactsidase activity(means+s.e.mean) and represent the results of at least four independent experiments. (b) Inhibition of IL-1b (1 ng ml71)-stimulated66TRE-Luc reporter gene by ¯uticasone propionate (FP) and dexamethasone (Dex). Results are expressed as relative light units/unit b-galactsidase activity (means+s.e.mean) and represent the results of at least four independent experiments.


In order to examine the potential mechanisms for thisrepression of GM-CSF release we examined the ability of thesedrugs to increase the expression of the b2-adrenoceptor and the

inhibitor of NF-kB-driven transcription, I-kB. The ability ofthese glucocorticoids to induce the b2-adrenoceptor was foundto correlate with GR/GRE binding and occurred at 10 ± 100fold higher concentrations than the inhibition of GM-CSF

release (see Table 1). These results are similar to those for¯uticasone and dexamethasone suppression of TNFa-inducedE-selectin expression (Ray et al., 1997). No induction of I-kBa,was seen in these cells by any steroid at concentrations up to1 mM for time periods up to 24 h. Furthermore, the ability ofIL-1b to cause I-kBa or I-kBb degradation and subsequent

induction was not a�ected by steroids. This con®rms resultsobtained in several other cell types where no e�ect ofdexamethasone was observed (Heck et al., 1997; Brostjan et

al., 1996; Ray et al., 1997) and suggest that induction of genetranscription by the activated glucocorticoid receptor was notrequired for inhibition of GM-CSF release.

The possibility of a direct interaction between activated GR

and NF-kB in the suppression of NF-kB activated genetranscription was indicated by the ability of these glucocorti-coids to inhibit a reporter gene construct containing three kBsites alone and was con®rmed by immunoprecipitationexperiments. This interaction is likely to occur through aleucine charged domain in the p65 subunit (Heery et al., 1997).

The inhibition of luciferase activity correlated with inhibitionof GM-CSF release suggesting that this is indeed an importantmechanism in regulating GR actions in these cells. However

only a 50 ± 60% inhibition of luciferase activity was seen atconcentrations of steroid at which GM-CSF release wascompletely inhibited. This suggests that although repressionof gene transcription plays a major role in the suppression of

GM-CSF release in these cells other post-transcriptional eventsmay also be important in glucocorticoid-repression of GM-CSF release.

The data presented here suggests that the anti-in¯ammatoryproperties of a range of glucocorticoids relate to their ability totransrepress rather than transactivate genes. Furthermore, the

results seen with RU486 suggests that repression of NF-kBand AP-1 activity by GR does not require the transactivationfunction of GR. The results also suggest that transrepressionoccurs at approximately 10 fold lower concentrations than that

required for transactivation of genes such as the b2-receptor.The ability of the more modern inhaled glucocorticoids¯uticasone propionate and budesonide, but not dexametha-

sone, to inhibit NF-kB activity appears to correlate moreclosely with GM-CSF release than the ability to inhibit AP-1activity. This suggests that NF-kB may be a more important

target for glucocorticoid actions, at least in the regulation ofin¯ammatory genes in A549 cells, than AP-1. However, in thecontext of other diseases and other cells AP-1 may be a more

important target. The importance of inhibiting both NF-kBand AP-1 within these cells may be relevant in the control ofin¯ammatory responses since these transcription factors areimportant for the expression of many genes and often act in

concert with each other (Stein et al., 1993).Similar results have been demonstrated for the interaction

between GR and AP-1 (Heck et al., 1994). In these studies

using overexpression vectors, a variety of GR mutants and achoice of glucocorticoid ligands, it has been possible todissociate the transrepressive and transactivation functions of

GR on AP-1 mediated reporter gene activities. Moreover,DNA binding and activation of glucocorticoid-regulatedpromoters require GR dimerisation, whereas AP-1, andprobably NF-kB, repression may be mediated by GR

monomers (Heck et al., 1994). Increased levels of either AP-1or NF-kB, raised by overexpression of cDNAs, also haveprofound e�ects on the ability of GR to inhibit transcription

driven by these genes. Expression of GR at levels that causemarked inhibition of either AP-1 or NF-kB alone fails torepress either factor when both are activated together(Scheinman et al., 1995b).

Recent results from several laboratories have providedsimilar but slightly di�erent results regarding the exactmechanisms of glucocorticoid suppression of gene transcrip-

tion. These di�erences may be due to the level of expression ofGR in each system since some depend upon over-expression ofGR, some occur in the presence of serum and others in the

absence of serum. We have recently found that the removal ofserum (or the use of charcoal-stripped serum) results inmarkedly up-regulated expression of GR at 48 h (Adcock et

al., unpublished observations).Studies of transcription factor interactions may have

therapeutic potential in the control of in¯ammatory disease.Glucocorticoids exert their anti-in¯ammatory e�ects largely by

interference with the ability of transcription factors that havebeen activated by in¯ammatory cytokines to induce transcrip-tion of in¯ammatory genes. This interaction between GR and

transcription factors may be either direct or through anintegrator molecule such as CBP or associated co-activatorssuch as p300/CBP associated protein (P/CAF), steroid-

receptor coactivator-1 (SRC-1) or p300/CBP co-integratorprotein (pCIP) (Janknecht & Hunter, 1996). Interaction ofthese factors with CBP allows interaction, modulation of

histone acetylation and subsequent activation of the basaltranscription initiation complex. Although not shown here forNF-kB, it is possible that this is the mechanism by whichglucocorticoids interfere with NF-kB-driven gene transcription

(Perkins et al., 1997; Sheppard et al., 1998). These interactionsmay a�ect transactivation by the pro-in¯ammatory transcrip-tion factor due to e�ects on DNA-binding, association with

the integrator molecules or activation of RNA polymerase IIto varying degrees. The exact contribution of each mechanismmay vary between cell types and depend upon the cell stimulus.

The di�erences in activity of di�erent glucocorticoids did notcorrelate solely with ligand binding but may re¯ect di�erentialinteraction of the same receptor with DNA or proteinsaccording to the activating ligand. Similar ligand-induced

di�erentiation of transrepression and transactivation activitieshas been reported previously for the interaction betweenretinoic acid and its receptors (RARa) (Yang Yen et al., 1991).

Thus, there is potential for the development of novelglucocorticoids with enhanced transrepressive and reducedtransactivation actions. Other drugs that regulate the activity

of speci®c transcription factors may also be developed in thefuture.

In conclusion, both trans-repression and trans-activation by

GR may be a�ected by the ligand in a manner that re¯ects notonly ligand a�nity but also the di�erential abilities of ligandsto in¯uence GR interaction with DNA, other transcriptionfactors, integrator molecules or the basal transcription

apparatus. The di�erentiation of these two activities (trans-repression and trans-activation) leads to the possibility of thedevelopment of more speci®c glucocorticoids. Indeed, these

results suggest that current drugs used in asthma therapy mayalready have been selected for an enhanced ability totransrepress rather than activate gene transcription.

This work was funded by the Medical Research Council (U.K.), TheNational Asthma Campaign, The European Union Biomed IIprogramme and Glaxo-Wellcome.


christinehurd

Highlight

christinehurd

Highlight

References

ADCOCK, I.M., GILBEY, T., GELDER, C.M., CHUNG, K.F. & BARNES,

P.J. (1996). Glucocorticoid receptor localization in normal andasthmatic lung. Am. J. Respir. Crit. Care Med., 154, 771 ± 782.

ADCOCK, I.M., LANE, S.J., BROWN, C.R., PETERS, M.J., LEE, T.H. &

BARNES, P.J. (1995). Di�erences in binding of glucocorticoidreceptor to DNA in steroid-resistant asthma. J. Immunol., 154,3500 ± 3505.

AUPHAN, N., DIDONATO, J.A., ROSETTE, C., HELMBERG, A. &

KARIN, M. (1995). Immunosuppression by glucocorticoids:inhibition of NF-kappa B activity through induction of I kappaB synthesis. Science, 270, 286 ± 290.

BARNES, P.J. (1995). Inhaled glucocorticoids for asthma. N. Engl. J.Med., 332, 868 ± 875.

BARNES, P.J. & ADCOCK, I. (1993). Anti-in¯ammatory actions ofsteroids: molecular mechanisms. Trends. Pharmacol. Sci., 14,436 ± 441.

BEATO,M., TRUSS, M. & CHAVEZ, S. (1996). Control of transcriptionby steroid hormones. Ann. N. Y. Acad. Sci., 784, 93 ± 123.

BRATTSAND, R. & LINDEN, M. (1996). Cytokine modulation byglucocorticoids: mechanisms and actions in cellular studies.Aliment. Pharmacol. Ther., 10 (Suppl 2), 81 ± 90.

BROSTJAN, C., ANRATHER, J., CSIZMADIA, V., STROKA, D.,

SOARES, M., BACH, F.H. & WINKLER, H. (1996). Glucocorti-coid-mediated repression of NFkappaB activity in endothelialcells does not involve induction of Ikappa Balpha synthesis. J.Biol. Chem., 271, 19612 ± 19616.

COLLINS, S., CARON, M.G. & LEFKOWITZ, R.J. (1988). Beta-adrenergic receptors in hamster smooth muscle cells aretranscriptionally regulated by glucocorticoids. J. Biol. Chem.,263, 9067 ± 9070.

DAVIDSON, F.F., LISTER, M.D. & DENNIS, E.A. (1990). Binding andinhibition studies on lipocortins using phosphatidylcholinevesicles and phospholipase A2 from snake venom, pancreas,and a macrophage-like cell line. J. Biol. Chem., 265, 5602 ± 5609.

FLOWER, R.J. & ROTHWELL, N.J. (1994). Lipocortin-1: cellularmechanisms and clinical relevance. Trends. Pharmacol. Sci., 15,71 ± 76.

HECK, S., BENDER, K., KULLMANN, M., GOTTLICHER, M.,

HERRLICH, P. & CATO, A.C. (1997). I-kappaB alpha independentdownregulation of NF-kappaB activity by glucocorticoidreceptor. EMBO J., 16, 4698 ± 4707.

HECK, S., KULLMANN, M., GAST, A., PONTA, H., RAHMSDORF,

H.J., HERRLICH, P. & CATO, A.C. (1994). A distinct modulatingdomain in glucocorticoid receptor monomers in the repression ofactivity of the transcription factor AP-1. EMBO J., 13, 4087 ±4095.

HEERY, D.M., KALKHOVEN, E., HOARE, S. & PARKER, M.G. (1997).A signature motif in transcriptional co-activators mediatesbinding to nuclear receptors. Nature, 387, 733 ± 736.

JANKNECHT, R. & HUNTER, T. (1996). Versatile molecular glue.Transcriptional control. Curr. Biol., 6, 951 ± 954.

KARIN, M. (1998). New twists in gene regulation by glucocorticoidreceptor: is DNA binding dispensable? Cell, 93, 487 ± 490.

KOCHETKOVA, M. & SHANNON, M.F. (1996). DNA triplexformation selectively inhibits granulocyte-macrophage colony-stimulating factor gene expression in human T cells. J. Biol.Chem., 271, 14438 ± 14444.

LEVINE, S.J. (1995). Bronchial epithelial cell-cytokine interactions inairway in¯ammation. J. Investig. Med., 43, 241 ± 249.

NEWTON, R., ADCOCK, I.M. & BARNES, P.J. (1996). Superinductionof NF-kappa B by actinomycin D and cycloheximide in epithelialcells. Biochem. Biophys. Res. Commun., 218, 518 ± 523.

PERKINS, N.D., FELZIEN, L.K., BETTS, J.C., LEUNG, K., BEACH,

D.H. & NABEL, G.J. (1997). Regulation of NF-kappaB by cyclin-dependent kinases associated with the p300 coactivator. Science,275, 523 ± 527.

RAY, K.P., FARROW, S., DALY, M., TALABOT, F. & SEARLE, N.

(1997). Induction of the E-selectin promoter by interleukin 1 andtumour necrosis factor alpha, and inhibition by glucocorticoids.Biochem. J., 328, 707 ± 715.

SCHEINMAN, R.I., COGSWELL, P.C., LOFQUIST, A.K. & BALDWIN,

JR A.S. (1995a). Role of transcriptional activation of I kappa Balpha in mediation of immunosuppression by glucocorticoids.Science, 270, 283 ± 286.

SCHEINMAN, R.I., GUALBERTO, A., JEWELL, C.M., CIDLOWSKI,

J.A. & BALDWIN, JR A.S. (1995b). Characterization of mechan-isms involved in transrepression of NF- kappa B by activatedglucocorticoid receptors. Mol. Cell Biol., 15, 943 ± 953.

SHEPPARD, K.A., PHELPS, K.M., WILLIAMS, A.J., THANOS, D.,

GLASS, C.K., ROSENFELD, M.G., GERRITSEN, M.E. & COLLINS,

T. (1998). Nuclear integration of glucocorticoid receptor andnuclear factor- kappaB signaling by CREB-binding protein andsteroid receptor coactivator-1. J. Biol. Chem., 273, 29291 ± 29294.

STEIN, B., BALDWIN, JR A.S., BALLARD, D.W., GREENE, W.C.,

ANGEL, P. & HERRLICH, P. (1993). Cross-coupling of the NF-kappa B p65 and Fos/Jun transcription factors producespotentiated biological function. EMBO J., 12, 3879 ± 3891.

STRAHLE, U., SCHMID, W. & SCHUTZ, G. (1988). Synergistic actionof the glucocorticoid receptor with transcription factors. EMBOJ., 7, 3389 ± 3395.

YANG YEN, H.F., ZHANG, X.K., GRAUPNER, G., TZUKERMAN, M.,

SAKAMOTO, B., KARIN, M. & PFAHL, M. (1991). Antagonismbetween retinoic acid receptors and AP-1: implications for tumorpromotion and in¯ammation. New Biol., 3, 1206 ± 1219.

ZABEL, U., HENKEL, T., SILVA, M.S. & BAEUERLE, P.A. (1993).Nuclear uptake control of NF-kappa B by MAD-3, an I kappa Bprotein present in the nucleus. EMBO J., 12, 201 ± 211.

(Received January 25, 1999Revised March 18, 1999

Accepted March 23, 1999)


sample literature report - personal.tcu.edupersonal.tcu.edu/jmontchamp/f2019 files for...

Documents